Psychedelics and Addiction: Alcohol, Tobacco, Opioids and Community Models
Psychedelic-Assisted Medicine: Science and Practice
Learning objectives for this lesson:
- Summarize the randomized evidence for LSD, psilocybin and ketamine in alcohol use disorder and for psilocybin in smoking cessation, explain why the psilocybin alcohol trials conflict, and state what a pilot trial can establish.
- Calculate differences in heavy drinking days, odds ratios, risk differences and numbers needed to treat from published trial data.
- Evaluate the proposed mechanisms of change in addiction and identify the kind of evidence behind each.
- Evaluate the observational evidence for ibogaine in opioid use disorder alongside its pharmacology, its cardiac risk, the case series of deaths and the limits of small safety studies.
- Explain why Canadian guidance advises against withdrawal management alone for opioid use disorder, what that means for single-session ibogaine treatment, and how public funding and regulation are shaping ibogaine’s development.
- Describe ayahuasca’s composition, contexts of use and legal status in Canada, including religious exemptions, and summarize the research on ayahuasca and substance use, including the 2013 study of ayahuasca-assisted treatment in a First Nations community in British Columbia.
- Identify the populations and designs of Roots to Thrive’s published evaluations and judge which claims about community models the evidence supports.
- Summarize the strength of the evidence for each psychedelic approach to addiction as of October 2026 and identify the recurring threats to the validity of its trials.
- Compare psychedelic trial results with benchmarks from standard treatment, converting an odds ratio to a number needed to treat at a chosen baseline risk, and describe where psychedelic-assisted therapy could sit beside standard care and harm reduction in Canada.
This course was developed by Dr. Kiffer G. Card.
Alcohol and Tobacco: LSD, Psilocybin and Ketamine Trials and Their Proposed Mechanisms
Introduction
A substance use disorder is a pattern of substance use that leads to clinically significant impairment or distress. Alcohol use disorder, for example, is diagnosed when a person meets at least 2 of 11 criteria over a year, such as continuing to drink despite harm (Wood et al., 2023). HSCI 180 Lesson 8 developed the neurobiology of addiction, and HSCI 402 develops the standard treatment of substance use disorders. This lesson asks what the research shows about psychedelic-assisted therapy for addiction and how it would sit beside existing treatments.
Alcohol use disorder was an early focus of psychedelic research in the 1950s and 1960s, at Weyburn and Hollywood Hospital among other places (Lesson 2), and it has drawn renewed attention in recent trials. This section moves from the randomized LSD trials of the 1960s to modern trials of psilocybin and ketamine, asking of each the questions in Lesson 1’s checklist: design, comparator, blinding, sample size and follow-up.
Learning Objectives
- 1.1 Summarize Krebs and Johansen’s (2012) meta-analysis of LSD trials for alcohol use disorder and explain what its fading effect at 12 months implies.
- 1.2 Describe the psilocybin trials in alcohol use disorder, including Bogenschutz et al. (2022) and a Swiss phase 2 trial, and give reasons their results conflict.
- 1.3 Describe the psilocybin smoking cessation studies, including the 2026 pilot trial against nicotine patch, and state what a pilot trial can establish.
- 1.4 Summarize two randomized trials of ketamine in alcohol use disorder, Dakwar et al. (2020) and the KARE trial (Grabski et al., 2022).
- 2.1 Calculate a difference in the percentage of heavy drinking days and the relative reduction it represents.
- 2.2 Calculate an odds ratio, risk difference and number needed to treat from published abstinence counts.
- 3.1 Evaluate proposed mechanisms of change and identify the kind of evidence behind each.
1.1 The LSD Trials and Their Meta-Analysis
Between 1966 and 1970, research groups in Canada and the United States randomized patients in alcohol treatment programs to a single dose of LSD or a control condition, and in 2012 two Norwegian researchers pooled the results.
1.1.1 What Krebs and Johansen Found
Teri Krebs and Pål-Ørjan Johansen (2012) identified six randomized controlled trials with 536 participants, of whom 325 received a full dose of LSD and 211 a control condition (low-dose LSD, d-amphetamine, ephedrine or no drug); almost all were men in inpatient alcohol treatment. A meta-analysis combines the results of several studies into one pooled estimate, weighting each study by its precision.
Across all six trials, the pooled odds ratio, which compares the odds of improvement in the two groups, was 1.96 (95 percent confidence interval 1.36 to 2.84) at the first reported follow-up. In the five trials with dichotomous outcomes, 185 of 315 LSD patients (59 percent) and 73 of 191 control patients (38 percent) had improved, and the pooled difference in the proportion improved was 16 percentage points, a number needed to treat (NNT, introduced in Lesson 8) of about six to seven: the authors gave six, while 1 ÷ 0.16 = 6.25 rounds up to 7 (Krebs & Johansen, 2012). In the three trials reporting maintained abstinence, LSD also showed a benefit at first follow-up.
Figure 10.1 shows that the benefit did not last: it was present at two to three months and at six months, and by 12 months the confidence interval included 1, the value of no difference.
1.1.2 How Much Weight the Trials Can Bear
Between-trial heterogeneity was negligible, so the trials pointed the same way. Two trials concealed allocation only until the LSD session, two were judged at high risk of bias from possible selective outcome reporting, and the four with active placebos faced the problem, explained in Lesson 5, that blinding rarely survives a full psychedelic dose (Krebs & Johansen, 2012). The meta-analysis shows a consistent short-term signal among men treated as inpatients in the 1960s, from which the modern psilocybin trials began.
1.2 Psilocybin for Alcohol Use Disorder
A 2025 systematic review judged psilocybin (Lesson 3) for alcohol use disorder to have the best evidence of efficacy among psychedelic treatments for substance use disorders (Piper et al., 2025), and a 2026 meta-analysis of six trials described that evidence as mixed, with preliminary signals of efficacy (Florineth et al., 2026; Section 4). This part describes four psilocybin studies in alcohol use disorder that illustrate how their designs and results differ.
1.2.1 From a Proof-of-Concept Study to a Randomized Trial
In a proof-of-concept study, Michael Bogenschutz and colleagues (2015) gave psilocybin in one or two supervised sessions to 10 volunteers with alcohol dependence who were receiving motivational enhancement therapy, a counselling approach that helps people resolve ambivalence about change. Abstinence increased significantly after the first session, and the authors called for controlled trials, since a single-group study cannot separate the drug from therapy, time or expectation.
Their randomized trial followed (Bogenschutz et al., 2022). At two United States centres, 95 adults with alcohol dependence received 12 weeks of motivational enhancement therapy and cognitive behavioural therapy and were randomized to two sessions of psilocybin (25 to 40 mg per 70 kg) or of diphenhydramine, an antihistamine used as an active placebo. Over the 32 weeks after the first dose, the percentage of heavy drinking days was 9.7 percent in the psilocybin group and 23.6 percent in the diphenhydramine group, a mean difference of 13.9 percentage points (95 percent confidence interval 3.0 to 24.7). In the final month, 23 of 48 participants given psilocybin (47.9 percent) and 11 of 45 given diphenhydramine (24.4 percent) were abstinent. There were no serious adverse events among participants who received psilocybin.
The authors listed the trial’s limits. Participants guessed their allocation correctly after 93.6 percent of first sessions and 94.7 percent of second sessions, so expectancy could have influenced the results. A biological check of self-reported drinking was available for only 53.8 percent of participants, enrolment stopped at 95 of a planned 180 when COVID-19 halted recruitment, and participants drank less heavily at screening than those in most medication trials (Bogenschutz et al., 2022); Section 4 returns to who was excluded. The box below explains the drinking outcomes these trials report.
How alcohol trials measure drinking
Alcohol trials, including those in this section, commonly use the timeline followback, an interview in which the participant reconstructs their drinking day by day on a calendar. From it, researchers compute outcomes such as the percentage of heavy drinking days (in Bogenschutz et al., 2022, days with five or more drinks for a man or four or more for a woman) and abstinence. Some trials check self-report with ethyl glucuronide (EtG), a breakdown product of alcohol found in urine or hair. Continuous outcomes are compared as mean differences, and yes-or-no outcomes as odds ratios, risk differences or numbers needed to treat.
1.2.2 Two European Trials after Withdrawal Treatment
Two European trials published in 2025 enrolled people after withdrawal treatment. In a Swiss phase 2 trial, Rieser and colleagues (2025) randomized participants to a single 25 mg dose of psilocybin or to placebo (mannitol), with five sessions of brief psychotherapy, and analyzed the 37 who completed the four-week follow-up. The groups did not differ in abstinence or alcohol use at four weeks or six months, although craving fell in both groups, and further after psilocybin. In a French inpatient feasibility trial, Luquiens and colleagues (2025) gave 30 adults with severe alcohol use disorder and depressive symptoms two sessions of 25 mg or 1 mg psilocybin added to standard care. At 12 weeks, 11 of 20 in the 25 mg group and 1 of 9 in the 1 mg group were abstinent.
Table 10.1 sets the four studies side by side. Their differences suggest reasons for the conflict: one dose or two, five sessions of therapy or twelve weeks, participants recently detoxified or still drinking, and samples from 10 (open-label) to 95. Research in this area is moving quickly: as of October 2026, a French phase 3 trial planned for 172 participants was registered as recruiting (ClinicalTrials.gov, NCT07638553, n.d.). In Canada, one of three psilocybin trials for which the federal government announced nearly $3 million on 29 June 2023 concerns alcohol use disorder (Canadian Institutes of Health Research, 2023).
Table 10.1. Psilocybin trials in alcohol use disorder, with their design, comparator and main result.
| Study | Design and participants | Psilocybin and therapy | Comparator | Main result |
|---|---|---|---|---|
| Bogenschutz et al., 2015 (United States) | Single-group, open-label; 10 adults | One or two sessions with motivational enhancement therapy | None | Abstinence increased after the first session; gains largely held to 36 weeks |
| Bogenschutz et al., 2022 (United States) | Double-blind randomized trial; 95 adults, 93 analyzed | Two sessions within 12 weeks of motivational and cognitive behavioural therapy | Diphenhydramine (active placebo) | Heavy drinking days 9.7% versus 23.6% over 32 weeks |
| Rieser et al., 2025 (Switzerland) | Double-blind randomized phase 2 trial after withdrawal treatment; 37 analyzed | One 25 mg dose with brief psychotherapy | Mannitol (inactive placebo) | No difference in abstinence or alcohol use at 4 weeks or 6 months |
| Luquiens et al., 2025 (France) | Double-blind randomized feasibility trial in inpatients with depressive symptoms; 30 adults | Two 25 mg sessions added to standard care | 1 mg psilocybin | Abstinent at 12 weeks: 11 of 20 versus 1 of 9 |
1.3 Psilocybin for Smoking Cessation
Smoking cessation can be checked biologically through exhaled carbon monoxide and cotinine, a breakdown product of nicotine. Two studies from Johns Hopkins University used these checks.
1.3.1 The Open-Label Pilot
Matthew Johnson and colleagues (2014) gave 15 people who smoked, with a mean of six previous quit attempts, two or three doses of psilocybin (20 and 30 mg per 70 kg) within a 15-week cognitive behavioural program. At six months, 12 (80 percent) had biologically confirmed seven-day point prevalence abstinence, meaning no smoking in the seven days before the visit, and 9 (60 percent) were abstinent at a mean of 30 months (Johnson et al., 2017). The authors noted that the open-label design does not allow definitive conclusions about efficacy.
1.3.2 The 2026 Pilot Randomized Trial
In a pilot randomized trial run from 2015 to 2023, the same group randomized 82 psychiatrically healthy adults who smoked to one dose of psilocybin (30 mg per 70 kg) on their target quit date or to 8 to 10 weeks of nicotine patch, with the same 13-week cognitive behavioural program, and everyone knew which treatment they received (Johnson et al., 2026). The primary outcome was prolonged abstinence, defined as no smoking after an initial 14-day grace period following the quit date and confirmed by exhaled carbon monoxide and urinary cotinine.
At six months, 17 of 42 participants given psilocybin (40.5 percent) and 4 of 40 given the patch (10.0 percent) showed prolonged abstinence, an odds ratio of 6.12 (95 percent confidence interval 1.99 to 23.26); seven-day point prevalence abstinence was 52.4 percent and 25.0 percent. No serious adverse events were attributed to either treatment. The authors acknowledged that expectancy could have contributed, and they argued that an unblinded comparison was reasonable given how poorly double-blind psychedelic trials preserve the blind. The wide interval reflects that only four people in the patch group succeeded. Worked Example 10.2, later in this section, recomputes the main result.
1.4 Ketamine for Alcohol Use Disorder
Ketamine, the dissociative anaesthetic that Lessons 3 and 4 described, is an established medicine, so trials in addiction test a new use of a familiar drug, given by intravenous infusion.
In a pilot trial, Dakwar and colleagues (2020) gave 40 adults with alcohol dependence in motivational enhancement therapy one infusion of ketamine or of midazolam, a sedative used as an active control; ketamine increased abstinence and delayed relapse over the next 21 days. The larger KARE trial in England (Grabski et al., 2022) randomized 96 people with severe alcohol use disorder to three weekly ketamine or saline infusions, combined with mindfulness-based relapse prevention therapy or with alcohol education. At six months, the ketamine groups had more days abstinent (mean difference 10.1 percentage points, 95 percent confidence interval 1.1 to 19.0), the largest difference was between ketamine with therapy and saline with education, and relapse rates did not differ. A phase 3 trial (MORE-KARE), planned for 280 participants, began recruiting in the United Kingdom in June 2024; in October 2026 its registry record listed it as recruiting, with completion planned for November 2027 (ISRCTN85955128; ISRCTN registry, 2026). Table 10.2 summarizes the smoking and ketamine trials.
Table 10.2. Psilocybin trials in smoking cessation and ketamine trials in alcohol use disorder, with their design and main result.
| Study | Design | Participants | Comparator | Main result |
|---|---|---|---|---|
| Johnson et al., 2014 and 2017 | Open-label pilot | 15 adults who smoked | None | 80% abstinent at 6 months, 67% at 12 months, 60% at a mean of 30 months |
| Johnson et al., 2026 | Unblinded pilot randomized trial | 82 adults who smoked | Nicotine patch; both arms received CBT | Prolonged abstinence 40.5% versus 10.0% at 6 months |
| Dakwar et al., 2020 | Randomized pilot trial | 40 adults with alcohol dependence | Midazolam infusion | More abstinence and later relapse over 21 days |
| Grabski et al., 2022 (KARE) | Randomized, placebo-controlled phase 2, two-by-two | 96 adults with severe alcohol use disorder | Saline infusion; therapy or education | Days abstinent 10.1 points higher at 6 months; relapse unchanged |
2.1 Calculating a Difference in Heavy Drinking Days
Worked Example 10.1 turns the Bogenschutz et al. (2022) result, described under 1.2 Psilocybin for Alcohol Use Disorder, into heavy drinking days, using the formula below.
Worked Example 10.1: Heavy drinking days in Bogenschutz et al. (2022)
The control percentage was 23.57 and the psilocybin percentage 9.71, so the difference is 23.57 − 9.71 = 13.86 percentage points. The relative reduction is 13.86 ÷ 23.57 = 0.588, or 58.8 percent. The follow-up window ran from week 5 to week 36, which is 32 weeks or 224 days, so the difference corresponds to 13.86 ÷ 100 × 224 = 31.0 fewer heavy drinking days per person over the period.
Both groups improved from screening, when heavy drinking days were 56.5 percent in the psilocybin group and 48.6 percent in the control group. The control group’s improvement reflects therapy, trial attention and natural change, which is why a between-group comparison is needed.
Use Calculator 10.1 to repeat the worked example and to see how the same difference means different things at different control levels.
The hypothetical presets show that a five-point difference is a 50 percent reduction from a control level of 10 percent and a 10 percent reduction from 50 percent.
2.2 Odds Ratios, Risk Differences and Numbers Needed to Treat
The smoking trial described under 1.3 Psilocybin for Smoking Cessation reported its main result as an odds ratio. Worked Example 10.2 recomputes it from the published counts, using the formula below, and adds the risk difference and number needed to treat.
Worked Example 10.2: Prolonged abstinence in Johnson et al. (2026)
In the psilocybin group, 17 of 42 were abstinent and 25 were not, so the odds were 17 ÷ 25 = 0.680. In the patch group, 4 of 40 were abstinent and 36 were not, so the odds were 4 ÷ 36 = 0.111. The odds ratio is 0.680 ÷ 0.111 = 6.12, as published. The proportions were 0.405 and 0.100, so the risk difference is 0.305 and the number needed to treat is 1 ÷ 0.305 = 3.3.
The risk ratio is 0.405 ÷ 0.100 = 4.05. The odds ratio is larger because abstinence was common in the psilocybin group, and odds exaggerate a ratio when the outcome is common: “six times the odds” of quitting meant about four times as many people quit.
Calculator 10.2 repeats the worked example and adds counts from the alcohol trials; its interval uses Woolf’s approximation and differs somewhat from the published one.
The Krebs preset adds the five trials’ counts and gives an odds ratio of 2.30, larger than the pooled 1.96, because a meta-analysis compares groups within each trial before combining them. The Luquiens preset shows how one success in a control group of nine yields an interval from about 1 to more than 90.
A hypothetical news article states that “one dose of psilocybin makes people six times more likely to quit smoking than the patch.” Using Worked Example 10.2, rewrite it in two sentences that report the risk ratio, the design (an unblinded pilot trial of 82 adults) and the follow-up period, and name one feature that a larger trial would need to change.
3.1 Proposed Mechanisms of Change
Lessons 4 and 5 described the brain mechanisms and the experience. The proposed mechanisms of change in addiction rest on different kinds of evidence.
One proposal links benefit to the acute experience, and a 2022 systematic review found that mystical-type experience was associated with better outcomes in 10 of 12 studies, most of them limited by small samples (Ko et al., 2022). In the smoking pilot, the 12 participants abstinent at six months had scored higher on mystical-type experience than the three still smoking, while overall drug intensity did not differ (Garcia-Romeu et al., 2014). In the alcohol proof-of-concept study, first-session intensity predicted less drinking, less craving and more abstinence self-efficacy, a person’s confidence that they can avoid drinking (Bogenschutz et al., 2015). These are correlations in small samples, and people already changing may respond more strongly. For ketamine, a secondary analysis of KARE found that no measured psychoactive effect mediated the effect on days abstinent (Bloy et al., 2026). In people, these mechanisms remain largely speculative. The flip cards below summarize four proposals and their evidence.
Summary and Transition
The evidence for psychedelics in alcohol and tobacco use disorders rests mainly on small early trials, and in alcohol use disorder their results are mixed. LSD showed a benefit that faded by 12 months, psilocybin lowered heavy drinking days by 13.9 percentage points in a 2022 trial while a Swiss trial found no difference, and a 2026 meta-analysis described the psilocybin evidence as mixed and preliminary. An unblinded pilot found prolonged abstinence from smoking in 40.5 percent with psilocybin and 10.0 percent with nicotine patch, and two ketamine trials found more abstinence. The proposed mechanisms rest mainly on correlations in small studies and on animal research, and in people they remain largely speculative. Section 2 turns to ibogaine and opioid use disorder, where reviews describe the evidence as mostly observational and the risks are more serious.
Key Takeaways
- 1.1 Krebs and Johansen (2012) pooled six randomized trials from 1966 to 1970 with 536 participants, almost all men treated as inpatients, and found an odds ratio for improvement of 1.96 at first follow-up. The benefit was present at two to three months and at six months, but by 12 months the confidence interval included 1, the value of no difference. The trials show a consistent short-term signal, with the caveat that blinding rarely survives a full psychedelic dose.
- 1.2 Bogenschutz et al. (2022) found 9.7 versus 23.6 percent heavy drinking days with psilocybin and an active placebo over 32 weeks, while participants guessed their allocation after 93.6 percent of first sessions. A Swiss phase 2 trial of a single 25 mg dose after withdrawal treatment found no difference from placebo, while a French feasibility trial found more abstinence with two 25 mg sessions. The number of doses, the amount of therapy, the stage of recovery and the sample size may explain the conflict, and a 2026 meta-analysis of six trials described the evidence for psilocybin in alcohol use disorder as mixed and preliminary.
- 1.3 An open-label pilot of 15 people who smoked reported 80 percent biologically confirmed abstinence at six months, which its design could not attribute to psilocybin. The 2026 unblinded pilot trial of 82 adults found prolonged abstinence in 40.5 percent with psilocybin and 10.0 percent with nicotine patch. A pilot trial can show feasibility and estimate an effect for a larger trial, but with no blinding and only four successes in the patch group it cannot settle efficacy.
- 1.4 In a pilot trial, one ketamine infusion increased abstinence over 21 days compared with midazolam (Dakwar et al., 2020). In the KARE trial, three ketamine infusions increased days abstinent by 10.1 percentage points at six months, with no difference in relapse, and the largest difference was between ketamine with therapy and saline with education. In October 2026 the registry record of a phase 3 trial listed it as recruiting.
- 2.1 The difference in the percentage of heavy drinking days is the control percentage minus the treatment percentage: in Bogenschutz et al. (2022), 23.57 minus 9.71 gives 13.86 percentage points, a relative reduction of 58.8 percent and about 31 fewer heavy drinking days per person over 224 days. The same absolute difference can be a large or a small relative change depending on the control level, and because both groups improved from screening, only the between-group comparison isolates the effect of psilocybin.
- 2.2 The odds ratio divides the odds of success in the treatment group by the odds in the control group, the risk difference subtracts the two proportions, and the number needed to treat is one divided by the risk difference. In the 2026 smoking trial these were 6.12, 0.305 and 3.3, while the risk ratio was 4.05. Odds exaggerate a ratio when the outcome is common, and intervals widen sharply when a group has few successes.
- 3.1 One proposal links benefit to mystical-type or meaningful experience, but its evidence consists of correlations in small samples, and a secondary analysis of KARE found that no measured psychoactive effect mediated ketamine’s benefit. Other proposals invoke motivation, craving and self-efficacy, neuroplasticity, which rests mainly on animal research, and the therapy itself, since control groups improved substantially in the trials reviewed here.
Reflection
A hypothetical journalist writes that “psilocybin works for addiction.” Three trials bear on the claim. (a) Bogenschutz et al. (2022) randomized 95 adults with alcohol dependence to psilocybin or diphenhydramine, both with 12 weeks of psychotherapy; heavy drinking days over 32 weeks were 9.7 percent and 23.6 percent, and participants guessed their allocation after 93.6 percent of first sessions. (b) Rieser et al. (2025) randomized people who had recently completed alcohol withdrawal treatment to one 25 mg dose of psilocybin or placebo with brief psychotherapy and, among 37 analyzed, found no difference in abstinence or alcohol use at four weeks or six months. (c) Johnson et al. (2026), an unblinded pilot trial of 82 adults who smoked, found prolonged abstinence at six months in 40.5 percent with psilocybin and 10.0 percent with nicotine patch. In 150 to 200 words, write a reply to the journalist that says what the evidence supports, where it conflicts and two design features that limit confidence.
Minimum 20 characters required.
Question 1: Krebs and Johansen (2012) pooled six randomized trials of LSD for alcohol use disorder conducted between 1966 and 1970. Which statement best describes how the benefit changed over follow-up?
Question 2: In Bogenschutz et al. (2022), the percentage of heavy drinking days over 32 weeks was 9.7 percent with psilocybin and 23.6 percent with diphenhydramine. Which interpretation is most accurate?
Question 3: In the 2026 pilot trial, 17 of 42 participants given psilocybin and 4 of 40 given nicotine patch showed prolonged abstinence at six months. Which statement is correct?
Question 4: Which finding comes from the KARE trial of ketamine in alcohol use disorder (Grabski et al., 2022)?
Ibogaine and Opioid Use Disorder: Observational Evidence, Cardiac Risk and Drug Development
Content note
This section discusses deaths associated with ibogaine and the risk of death from opioid overdose. If you would like support, the 9-8-8 Suicide Crisis Helpline can be reached by call or text at 9-8-8, and the Hope for Wellness Help Line for Indigenous people is available at 1-855-242-3310.
Introduction
Opioid use disorder is a substance use disorder involving opioids such as heroin, fentanyl or oxycodone. In Canada, the 2018 national guideline recommends opioid agonist treatment, in which a long-acting opioid medication replaces the unpredictable supply; buprenorphine-naloxone is the preferred first-line option and methadone is the alternative (Bruneau et al., 2018). HSCI 402 develops these treatments. Agonist treatment is usually long term, and some people look for a treatment that would end opioid use altogether. Ibogaine is used in medical and non-medical settings for that purpose (Brown & Alper, 2018). This section examines the evidence that ibogaine helps people with opioid use disorder, the evidence that it can stop the heart, and the policy decisions that now push it toward drug development.
Learning Objectives
- 4.1 Describe ibogaine’s origins in Bwiti religious practice and the features of its pharmacology that matter for treating opioid use disorder.
- 4.2 Summarize the observational studies of ibogaine for opioid use disorder and a randomized safety trial of noribogaine, and identify what each design can establish.
- 4.3 Explain how ibogaine and noribogaine prolong the QT interval, and summarize the case series of deaths after ibogaine use.
- 4.4 Calculate an upper bound for the rate of a rare adverse event from a study that observed few or none.
- 4.5 Evaluate the evidence for giving magnesium with ibogaine and explain why an uncontrolled study of 30 people cannot establish safety.
- 5.1 Explain why Canadian guidance advises against withdrawal management alone for opioid use disorder and what that means for single-session ibogaine treatment.
- 5.2 Describe state-funded ibogaine development in Texas, the 2026 United States executive order and ibogaine’s legal status in Canada.
4.1 Ibogaine’s Origins and Pharmacology
Ibogaine is an alkaloid from the root of Tabernanthe iboga, a shrub of the West African rainforest. Peoples of West Central Africa who follow the Bwiti religion have used iboga as a sacrament for centuries, including in initiation rites (Mosca et al., 2023), and Lesson 2 described that tradition and what outside attention has meant for it. As a medicine, ibogaine is usually given by mouth as ibogaine hydrochloride. The liver converts it, mainly through the enzyme CYP2D6, into noribogaine, an active metabolite that stays in the body far longer than ibogaine, with a half-life of about 24 to 30 hours in one trial (Glue et al., 2016; Brunt, 2026); other pharmacokinetic studies report 28 to 49 hours, the range Lesson 6 uses (Koenig & Hilber, 2015). Lesson 3 described its mixed receptor actions.
4.2 The Observational Evidence
A 2023 systematic review found that most studies of ibogaine or noribogaine for substance use disorders were case reports, case series or observational studies, with two double-blind placebo-controlled trials (Mosca et al., 2023). The studies described here illustrate that evidence; each is informative, and each leaves open different questions.
4.2.1 Four Studies of Ibogaine and Noribogaine
In a case series, Mash and colleagues (2018) reviewed clinical results from 191 volunteers seeking to stop using opioids or cocaine during medically monitored inpatient treatment. Ibogaine reduced opioid withdrawal symptoms and drug craving, and the authors reported no significant adverse events in the dose range used. Brown and Alper (2018) followed 30 people with opioid dependence treated at a clinic in Mexico (Brown et al., 2019), who had had an average of 3.1 previous episodes of treatment. Scores on the Subjective Opioid Withdrawal Scale fell from 31.0 before treatment to 14.0 about three days afterward, 15 participants (50 percent) reported no opioid use in the 30 days before the one-month visit, and scores on the Addiction Severity Index drug use scale remained better than baseline to 12 months.
In New Zealand, where ibogaine could be legally prescribed, Noller, Frampton and Yazar-Klosinski (2018) followed 14 people (half of them women) for 12 months after a single treatment by one of two providers. The eight participants who completed every interview showed a significant reduction in the Addiction Severity Index drug use score, depression scores fell, and withdrawal scores across all 14 participants fell significantly just after treatment. One participant enrolled in the study died during treatment.
A randomized trial in opioid use disorder tested noribogaine. Glue and colleagues (2016) gave single doses of noribogaine (60, 120 or 180 mg) or placebo to 27 patients who wished to stop methadone treatment and had been switched to morphine for a week. Noribogaine was well tolerated and showed a trend toward lower withdrawal scores that did not reach statistical significance. It also lengthened the QT interval in proportion to its concentration, with the largest mean increases of about 16, 28 and 42 milliseconds at the three doses. The 2023 systematic review, with searches to November 2021, selected 31 articles and concluded that ibogaine shows some efficacy but that its cardiotoxicity and mortality are worrying (Mosca et al., 2023). Table 10.3 compares the studies.
Table 10.3. Studies of ibogaine and noribogaine in opioid use disorder and related conditions, with their design, participants and main findings.
| Study | Design | Participants and setting | Main findings | Main limits |
|---|---|---|---|---|
| Mash et al., 2018 | Open-label case series | 191 people stopping opioids or cocaine, inpatient | Less withdrawal and craving; no significant adverse events reported | No comparison group; follow-up of one month where available |
| Brown & Alper, 2018 | Prospective observational | 30 people with opioid dependence, Mexico | Withdrawal score 31.0 to 14.0; 50% no opioid use at one month | No comparison group; self-report |
| Noller et al., 2018 | Prospective observational | 14 people with opioid dependence, New Zealand | Lower drug use scores at 12 months in the 8 who completed every interview; one death during treatment | Small; 6 of 14 incomplete |
| Glue et al., 2016 | Randomized, double-blind, placebo-controlled | 27 people stopping methadone | Non-significant trend toward less withdrawal; QT lengthened with dose | Single doses; designed to test safety |
| Cherian et al., 2024 | Prospective observational, open-label | 30 male veterans with traumatic brain injury, Mexico | Large improvements in function and symptoms at one month; no serious adverse events | No comparison group; not opioid use disorder |
4.2.2 What Uncontrolled Studies Can Show
A study without a comparison group can show that something is feasible and estimate how people fared, and it can detect large and unexpected harms. It cannot show that ibogaine caused the improvement. Withdrawal symptoms fade over days in any case, people often enter treatment when their problems are at their worst and then improve (a pattern called regression to the mean), and people who seek out and pay for ibogaine may be unusually motivated. Loss to follow-up adds uncertainty, because the participants who stay in touch may be those who are doing well. Interactive 10.1 asks you to classify each study’s design and read what it can establish.
For each study, choose the design that best describes it, then check your answers to read what the design can and cannot establish.
The interactive makes a pattern visible: the studies that measured benefit had no comparison group, and the randomized trial among them was designed to test safety. That imbalance matters most when the possible harm is death.
4.3 Cardiac Risk and Deaths
Lesson 6 introduced ibogaine’s effect on the heart as part of the physical risks of psychedelics. This part examines the evidence in more detail, because it decides whether any benefit could be worth pursuing.
4.3.1 The QT Interval
The QT interval is the time on an electrocardiogram from the start of the ventricles’ electrical activation to the end of their recovery, or repolarization; corrected for heart rate, it is written QTc. When repolarization is slowed, the QT interval lengthens and the heart becomes vulnerable to torsades de pointes, a ventricular arrhythmia that can be fatal. Ibogaine and noribogaine block hERG potassium channels in heart muscle cells, which carry a current that helps end each heartbeat’s electrical cycle, and this blockade is the main mechanism by which they prolong the QT interval (Brunt, 2026). Some deaths have occurred hours to days after ingestion, which the same review links to the persistence of noribogaine. Figure 10.2 traces the pathway and the factors that the evidence links to greater risk.
4.3.2 The Case Series of Deaths
Kenneth Alper and colleagues (2012) reviewed every available autopsy, toxicology and investigative report for all known deaths outside West Central Africa that followed ibogaine use between 1990 and 2008. Nineteen people, 15 men and 4 women aged 24 to 54, died between 1.5 and 76 hours after taking ibogaine. In 12 of the 14 deaths with adequate post-mortem information, pre-existing medical conditions, mainly cardiovascular, or other drugs taken at the same time explained or contributed to the death. The authors also identified seizures during withdrawal from alcohol or benzodiazepines and the uninformed use of traditional plant preparations of ibogaine as risk factors. A 2026 review added that case reports describe dangerous arrhythmias after typical treatment doses in people without known heart disease, and that people who metabolize ibogaine slowly through CYP2D6 may face higher risk; it recommended that any ibogaine treatment take place only under controlled medical supervision, with CYP2D6 genotyping and rigorous cardiovascular monitoring (Brunt, 2026).
A case series counts deaths without a denominator, and the cohort studies in Table 10.3 treated too few people to detect rare events. A preprint posted in June 2026, which had not been peer reviewed, pooled records from 11 international clinics and reported six deaths within 72 hours among 10,382 patients treated for opioid use disorder and none among 8,689 patients treated for conditions other than substance use disorders; the authors describe the rate as a lower bound, because clinics took part voluntarily and deaths were not independently adjudicated (Arns et al., 2026).
4.4 Upper Bounds for Rare Adverse Events
Worked Example 10.3 shows how much uncertainty a small study leaves.
Worked Example 10.3: What zero serious events in 30 people can tell us
The MISTIC study, described under 4.5 Magnesium and the MISTIC Study, reported no serious adverse events among 30 participants (Cherian et al., 2024). By the rule of three, the upper 95 percent limit for the true rate is 3 ÷ 30 = 0.10, or 10 percent. The exact one-sided limit is 1 − 0.051/30 = 0.095, or 9.5 percent.
The study is therefore compatible with a serious event rate as high as about 1 in 10. The same reasoning applied to the 191 patients in Mash et al. (2018) gives an upper limit of about 1.6 percent, which still cannot rule out a fatal arrhythmia in 1 of every 100 treatments.
Use Calculator 10.3 to repeat the worked example and to see how many participants a study would need before its upper limit fell below 1 in 1,000.
Excluding a risk of 1 in 1,000 requires roughly 3,000 participants with no events, ten times the 292 participants in Table 10.3 combined. Safety for rare fatal events therefore has to come from large registries or regulated programs that count every person treated.
4.5 Magnesium and the MISTIC Study
Researchers have tried to reduce the cardiac risk while keeping any benefit. One approach gives magnesium alongside ibogaine, because magnesium has been shown to shorten the QT interval and to protect against QT prolongation caused by other medications (Cherian et al., 2024). A 2024 study of that approach, described in Case 10.1, did not involve opioid use disorder.
Cherian and colleagues (2024) at Stanford University studied 30 male veterans of United States Special Operations Forces with mostly mild traumatic brain injury who had arranged, on their own and with a grant from a non-profit organization, to receive ibogaine at a clinic in Mexico, where its use is not restricted. The Stanford team assessed the veterans before and after and played no role in giving the drug. At the clinic, staff gave an intravenous infusion of magnesium sulfate before a total oral ibogaine dose below 14 mg per kg and again about 12 hours later, and monitored heart rhythm continuously.
The authors reported no unexpected or serious adverse events and no instances of slowed or racing heart rate, clinically meaningful QT prolongation that was visible on monitoring, or unstable blood pressure. Disability, post-traumatic stress, depression and anxiety scores fell steeply by one month. The authors stated that the study was not a randomized controlled trial, that expectancy and the clinic’s other activities could explain part of the benefit, and that controlled trials are needed.
Case 10.1 shows what an open-label study can contribute and where it stops. It describes a protocol that may reduce cardiac risk, and its participants were screened and monitored throughout. It did not compare ibogaine with and without magnesium, so it cannot show that magnesium was protective, and Worked Example 10.3 showed that 30 people without serious events leave a wide range of possible risk. Its participants did not have opioid use disorder, and the course draws no conclusion from it about clinics that operate outside regulation, where screening, monitoring and emergency care vary and no regulator counts outcomes.
5.1 Withdrawal Management Alone and the Risk of Overdose
Ibogaine is usually offered as a single treatment that interrupts opioid withdrawal. That framing raises a problem Canadian guidance has already addressed for conventional detoxification, which the box below summarizes.
Why withdrawal management alone is not recommended
Canada’s 2018 national guideline for opioid use disorder recommends that withdrawal management alone be avoided, because it has been associated with higher rates of relapse, illness and death; it cites the loss of tolerance after withdrawal, combined with high rates of relapse, as a reason that this practice increases the risk of overdose death (Bruneau et al., 2018). A person who has stopped opioids for days or weeks can no longer tolerate the dose they used before, so a return to use carries a higher risk of overdose.
The guideline applies directly to how ibogaine is used. If ibogaine works mainly by ending withdrawal in one session, people treated with it face the same period of lost tolerance as people leaving any detoxification program, and in the Mexican cohort only half of the participants reported no opioid use in the month after treatment (Brown & Alper, 2018). A study of ibogaine that meets current standards would therefore need to connect participants to ongoing care, including opioid agonist treatment for those who want it, and to count overdoses after treatment as an outcome. Section 4 returns to how psychedelic approaches would sit beside agonist treatment.
5.2 State-Funded Drug Development and Regulation
Ibogaine has moved from the margins of research toward public funding in the United States, while its legal status differs widely across countries. The accordion below summarizes the main developments, each dated, and Lesson 11 develops access and regulation in full.
As of October 2026, ibogaine was listed in Schedule I under United States federal law (Drug Enforcement Administration, 2026), a category defined as having no currently accepted medical use and a high potential for misuse (Mosca et al., 2023). Any medical use therefore requires an approved clinical trial or another federal pathway.
Governor Greg Abbott signed a law committing $50 million in state funds, to be matched by private investment such as a drug developer, toward clinical trials intended to bring ibogaine to approval by the Food and Drug Administration (FDA) for opioid use disorder and other conditions. If the FDA approves ibogaine, the state would receive a percentage of revenue from the drug (High Plains Public Radio, 2025).
The order directs the FDA and the Drug Enforcement Administration to create a pathway for eligible patients to access psychedelic drugs, “including ibogaine compounds,” under the Right to Try Act, and directs the Department of Health and Human Services, through the Advanced Research Projects Agency for Health (ARPA-H), to allocate at least $50 million to support state psychedelic programs (The White House, 2026). The order does not itself reschedule ibogaine; it asks for a review so that drugs approved after phase 3 trials could be rescheduled quickly. In 2026, ARPA-H invited proposals for phase 1 and phase 2 trials of ibogaine for opioid use disorder, noting that no federally authorized trial of that kind was then active in the United States (ARPA-H, 2026).
As of October 2026, ibogaine was not listed in the schedules of the Controlled Drugs and Substances Act (Justice Laws Website, 2026), and on 19 May 2017 Health Canada added ibogaine and its salts, derivatives and analogues to the Prescription Drug List (Health Canada, 2017). Lesson 11 describes the routes by which unapproved drugs can be accessed.
Ibogaine could be legally prescribed in New Zealand, which made the Noller et al. (2018) study possible. The 2023 review notes that ibogaine is also given in clinics in countries where its use is unregulated, such as Mexico, Thailand and the Netherlands (Mosca et al., 2023). None of the evidence in this section establishes that treatment in such settings is safe.
Public funding of drug development is a different decision from endorsing current practice. The Texas program pays for trials whose purpose is to find out whether ibogaine works and how safely it can be given, which is the evidence that the studies in this section could not provide.
Write three sentences for a health policy brief describing what a trial funded under Texas Senate Bill 2308 would need to show before ibogaine could be considered for opioid use disorder. Your sentences should name a comparison group, an outcome that captures overdose risk after treatment, and a safety requirement drawn from Figure 10.2 and Worked Example 10.3.
Summary and Transition
This section showed that the evidence for ibogaine in opioid use disorder comes mostly from case series and small cohorts without comparison groups, as a 2023 systematic review found; these studies report less withdrawal and less drug use and cannot show that ibogaine caused the change. Ibogaine and noribogaine block hERG channels and prolong the QT interval, a case series documented 19 deaths between 1990 and 2008, and small studies with no serious events still leave wide uncertainty about rare fatal events. Magnesium co-administration is a promising protocol, and the MISTIC study could not test whether it protects, because every participant received it. Canada’s 2018 guideline advises against withdrawal management alone because of overdose after lost tolerance, which applies to single-session ibogaine treatment. Public funding in Texas and from the United States government is intended to support trials that could provide controlled evidence, so this picture may change. Section 3 turns to ayahuasca and to community and group models of psychedelic care, including two in British Columbia.
Key Takeaways
- 4.1 Ibogaine is an alkaloid from the root of Tabernanthe iboga, used for centuries as a sacrament in the Bwiti religion of West Central Africa. The liver converts it, mainly through CYP2D6, into noribogaine, an active metabolite with a half-life of about 24 to 49 hours across studies, which stays in the body far longer than ibogaine itself.
- 4.2 Evidence for ibogaine in opioid use disorder comes mostly from studies without comparison groups, as a 2023 systematic review found; those described here, a case series of 191 inpatients and cohorts of 30 and 14 people, reported less withdrawal and less drug use. The one randomized trial described here, of noribogaine in 27 patients, was designed to test safety and showed a non-significant trend toward less withdrawal. Uncontrolled studies can show feasibility and detect large harms, but they cannot separate the drug from fading withdrawal, regression to the mean, motivation and loss to follow-up.
- 4.3 Ibogaine and noribogaine block hERG potassium channels in heart muscle cells, which slows repolarization, lengthens the QT interval and can lead to torsades de pointes and sudden death. A case series documented 19 deaths outside West Central Africa between 1990 and 2008, most with heart disease or other drugs contributing, and a 2026 review recommended treatment only under controlled medical supervision with CYP2D6 genotyping and cardiovascular monitoring.
- 4.4 When no events occur among n participants, the upper limit of a one-sided 95 percent confidence interval for the true rate is about 3 divided by n. Zero serious events among 30 people is compatible with a rate of about 1 in 10, and excluding a risk of 1 in 1,000 requires roughly 3,000 participants without events, so safety against rare fatal events has to come from large registries or regulated programs.
- 4.5 Magnesium can shorten the QT interval, and in the MISTIC study 30 veterans with traumatic brain injury who received magnesium with ibogaine under continuous monitoring had no serious adverse events. The study did not compare ibogaine with and without magnesium, did not involve opioid use disorder and was too small to establish safety, so the protocol is promising, and this study could not test whether it protects.
- 5.1 Canada's 2018 guideline for opioid use disorder advises against withdrawal management alone because lost tolerance, combined with high rates of relapse, raises the risk of overdose death. Single-session ibogaine treatment faces the same period of lost tolerance, so a study that meets current standards would need to connect participants to ongoing care, including agonist treatment, and count overdoses after treatment.
- 5.2 As of October 2026, ibogaine was a Schedule I drug in the United States. Texas committed $50 million in 2025, to be matched privately, toward trials intended to bring ibogaine to FDA approval, and Executive Order 14401 of April 2026 directed a Right to Try pathway and funding for state psychedelic programs. In Canada, as of October 2026, ibogaine was not in the schedules of the Controlled Drugs and Substances Act and had been on the Prescription Drug List since 2017; public funding of drug development is a different decision from endorsing current practice.
Reflection
A hypothetical community health board asks for a short evidence summary on ibogaine for opioid use disorder. You have these facts: Brown and Alper (2018) followed 30 people treated in Mexico, and 50 percent reported no opioid use in the month after treatment; Noller et al. (2018) followed 14 people in New Zealand for 12 months, with complete data for 8, and one participant died during treatment; Alper et al. (2012) reviewed 19 deaths after ibogaine use between 1990 and 2008, most with heart disease or other drugs contributing; ibogaine and noribogaine block hERG potassium channels and prolong the QT interval; a study of 30 veterans given magnesium with ibogaine reported no serious adverse events, which leaves an upper 95 percent limit of about 10 percent for the true rate; and the Canadian guideline for opioid use disorder advises against withdrawal management alone because lost tolerance raises the risk of overdose death. In 150 to 200 words, summarize what is known, what is unknown and what kind of study would be needed.
Minimum 20 characters required.
Question 1: Why can the studies by Brown and Alper (2018) and Noller et al. (2018) not show that ibogaine caused the reductions in opioid use they observed?
Question 2: What is the main mechanism by which ibogaine and noribogaine prolong the QT interval?
Question 3: The MISTIC study reported no serious adverse events among 30 veterans given magnesium with ibogaine. What can a reader conclude?
Question 4: Why does the Canadian national guideline for opioid use disorder bear on single-session ibogaine treatment?
Ayahuasca and Community Models: Ceremony, Churches and Group Care in British Columbia
Introduction
The trials in Sections 1 and 2 treated people one at a time in clinics. Many uses of psychedelics for healing take place in groups: in Indigenous ceremonies, in religious congregations and in group therapy programs. The course’s viewpoint holds that these traditions deserve respect and that the relationship and the setting are part of the treatment, which makes group models worth studying on their own terms. This section examines ayahuasca, a ceremonial medicine that addiction researchers have studied mainly through observational designs, and then a group model of ketamine-assisted therapy developed in British Columbia. HSCI 214 develops the history of colonialism, and HSCI 8XX develops Indigenous research ethics; this section cites Indigenous-authored principles where they bear on the evidence.
Learning Objectives
- 6.1 Describe the composition of ayahuasca, its Indigenous and religious contexts and its legal status in Canada.
- 6.2 Summarize the research on ayahuasca and substance use, including observational studies of church members, two systematic reviews and a placebo-controlled trial in depression.
- 6.3 Describe the 2013 observational study of ayahuasca-assisted treatment in a First Nations community in British Columbia, its findings and the limits of its design.
- 6.4 Explain what religious exemptions for ayahuasca churches in the United States and Canada authorize and what they do not establish.
- 7.1 Describe Roots to Thrive’s group-based ketamine-assisted therapy in Nanaimo and identify the populations and designs of its published evaluations.
- 7.2 Judge whether a claim about a community model is supported by a randomized comparison, by uncontrolled evidence only, or by no evidence.
6.1 Ayahuasca: Composition, Contexts and Legal Status
Ayahuasca is a tea made from two Amazonian plants, the vine Banisteriopsis caapi and the leaves of Psychotria viridis. Indigenous peoples in countries such as Brazil, Ecuador and Peru have used it for medicinal, spiritual and cultural purposes since before European contact, and in the 20th century it spread into syncretic religious practices adopted by non-Indigenous people (Tupper, 2008).
6.1.1 How the Two Plants Work Together
The leaves supply N,N-dimethyltryptamine (DMT), a classic psychedelic that acts at the serotonin 2A receptor. Taken by mouth on its own, DMT is broken down by the enzyme monoamine oxidase A before it can reach the brain. The vine supplies harmine, harmaline and tetrahydroharmine, which inhibit that enzyme and so allow DMT to act (Bouso et al., 2012). Lesson 3 described this interaction and the drug interactions that follow from it. As of October 2026, both DMT and harmaline were listed in Schedule III of Canada’s Controlled Drugs and Substances Act, and the Act treats any substance that contains a controlled substance as controlled (section 2(2); Justice Laws Website, 2026), so possessing ayahuasca without an exemption or other authorization is prohibited. Figure 10.3 summarizes the pharmacology and the legal position.
6.1.2 Three Settings of Use
Ayahuasca reaches people through settings with different purposes, rules and evidence. The tabs below describe three of them.
Ayahuasca has been used by Indigenous peoples of the Amazon for medicinal, spiritual and cultural purposes since pre-Columbian times (Tupper, 2008). Lesson 2 described these traditions and the colonial suppression of Indigenous ceremonies. An Indigenous-led group reported concerns from many Indigenous Nations about cultural appropriation, lack of recognition of the sacred cultural positioning of these medicines, exclusionary practices in research and practice, and the patenting of traditional medicines (Celidwen et al., 2023). The box on ethical principles of traditional Indigenous medicine, under 6.3 A Study in a First Nations Community, describes the principles it proposed.
Several Brazilian syncretic churches use ayahuasca as a sacrament, and they have expanded their activities to urban Brazil, Europe and North America. Members typically drink ayahuasca at least twice a month (Bouso et al., 2012). Religious exemptions, described under 6.4 Religious Exemptions for Ayahuasca Churches, have allowed some of these churches to practise legally outside Brazil.
Ayahuasca is also given in retreats and in research. The retreat study in a First Nations community (Case 10.2) and a hospital-based randomized trial in Brazil (Palhano-Fontes et al., 2019) are two examples with very different designs. The evidence in this section does not describe the safety of retreats outside research settings.
6.2 Research on Ayahuasca and Substance Use
A 2022 systematic review found that the research on ayahuasca and substance use disorders published from 2016 to 2020 consisted of preclinical and observational studies; it judged the findings promising, noted that observational designs cannot show cause, and called for randomized trials (Rodrigues et al., 2022). This part describes the main observational studies and two systematic reviews.
Fábregas and colleagues (2010) compared people who regularly took ayahuasca in religious rituals in Brazil with people who did not use ayahuasca, in a rural Amazonian setting (56 ritual participants and 56 controls) and in cities (71 ritual participants and 59 controls), using the Addiction Severity Index. The ritual participants scored lower on the alcohol use and psychiatric status scales in both settings, and the rural ritual participants had often used illicit drugs in the past and had stopped, except for cannabis. A year later, the differences remained in the rural group and were no longer significant in the urban group. A related study of 127 people who regularly took ayahuasca in rituals and 115 controls who took part in other religions found no evidence of psychological maladjustment or cognitive impairment among the ritual participants over a year (Bouso et al., 2012).
These comparisons cannot separate ayahuasca from membership in a religious community, and people who keep attending for years may differ from those who stop. An earlier systematic review found five animal studies and five observational human studies; four of the human studies reported reductions in dependence symptoms or substance use, and the authors concluded that controlled studies were needed (Nunes et al., 2016). The placebo-controlled trials of ayahuasca described here concern other conditions. In one, 29 patients with treatment-resistant depression received a single dose of ayahuasca or placebo, and response at day 7 was 64 percent and 27 percent (Palhano-Fontes et al., 2019); a pilot trial in 17 people with social anxiety disorder also used a placebo (dos Santos et al., 2021). Lesson 8 examines the depression trials of psilocybin, LSD and ketamine.
6.3 A Study in a First Nations Community
A 2013 study in British Columbia shows what a small community study can document. Case 10.2 describes it as its authors reported it.
Gerald Thomas, Philippe Lucas, N. Rielle Capler, Kenneth Tupper and Gina Martin reported a preliminary observational study of ayahuasca-assisted treatment for problematic substance use and stress delivered in a rural First Nations community in British Columbia. The program, called the “Working with Addiction and Stress” retreats, combined four days of group counselling with two expert-led ayahuasca ceremonies. The researchers collected data from 12 participants before the retreat and six months afterward on psychological and behavioural factors related to problematic substance use, and gathered qualitative accounts of participants’ experiences at six months.
Scores for hopefulness, empowerment, mindfulness and the meaning and outlook subscales of quality of life improved significantly. Self-reported use of alcohol, tobacco and cocaine declined, and the reduction in problematic cocaine use was statistically significant, while cannabis and opiate use did not decline. All participants reported positive and lasting changes. The authors concluded that this form of ayahuasca-assisted therapy “appears to be associated with” improvements in factors related to problematic substance use among “a rural aboriginal population” and that more rigorous research is warranted.
Case 10.2 shows careful reporting of a small study. Its authors described it as preliminary and observational, and its design has the limits of the other uncontrolled studies in this lesson: 12 participants, no comparison group, self-reported substance use and a program that combined counselling, ceremony and community setting, so no single element can be credited. Its contribution lies in documenting a program delivered in a First Nations community, with outcomes that included hope, empowerment and meaning as well as substance use. Questions about who designs such programs, who holds the data and how results are shared belong to Indigenous research ethics, which HSCI 8XX develops. The box below summarizes principles that Indigenous scholars and practitioners have proposed for Western psychedelic research.
Ethical principles of traditional Indigenous medicine
An Indigenous-led, globally represented group of practitioners, activists, scholars, lawyers and human rights defenders used a global Indigenous consensus process to identify eight interconnected ethical principles for Western psychedelic research and practice that involves traditional Indigenous medicines: Reverence, Respect, Responsibility, Relevance, Regulation, Reparation, Restoration and Reconciliation (Celidwen et al., 2023). The authors wrote in response to concerns from many Indigenous Nations about cultural appropriation, lack of recognition of the sacred cultural positioning of these medicines, exclusionary practices in research and practice, and the patenting of traditional medicines, and they noted that Indigenous voices and leadership had been largely absent from the field. For each principle they described concrete problems and suggested actions; under Respect, for example, they call for relationships with Indigenous knowledge holders, adherence to regional protocols and free, prior and informed consent. Lesson 12 examines these principles alongside Two-Eyed Seeing and reciprocity.
6.4 Religious Exemptions for Ayahuasca Churches
Because DMT has long been a controlled substance in Canada and the United States, churches that use ayahuasca as a sacrament have needed legal protection. The routes differ between the two countries, and neither route is a finding about therapy.
In the United States, members of the União do Vegetal (UDV) receive communion by drinking hoasca, a tea brewed from Amazonian plants that contains DMT. After customs inspectors seized a shipment, the church sued under the Religious Freedom Restoration Act. On 21 February 2006, in Gonzales v. O Centro Espírita Beneficente União do Vegetal, the Supreme Court held that the government had failed to show, at the preliminary injunction stage, a compelling interest in barring the church’s sacramental use (Supreme Court of the United States, 2006), a decision that opened the way for the church to import, distribute and drink hoasca (Bullis, 2008). In Canada, the Minister of Health may exempt a person or group from the Controlled Drugs and Substances Act under section 56, and Lesson 2 described the exemptions that Canadian ayahuasca churches obtained through that route, the first in June 2017.
A religious exemption protects a congregation’s sacramental use, which is a different activity from offering ayahuasca as a health treatment, and it is not evidence that ayahuasca treats addiction. Studies of long-term ritual participants, like those under 6.2 Research on Ayahuasca and Substance Use, describe people who have chosen and stayed in a religious community, which limits what they can say about people who seek treatment.
7.1 Group-Based Ketamine-Assisted Therapy in British Columbia
The second community model in this section uses a legal medicine in a group format. Roots to Thrive in Nanaimo is one of the course’s running Canadian examples, and Lesson 7 described its therapeutic process. This part examines its evidence.
7.1.1 The Roots to Thrive Model
Roots to Thrive began in 2018 as a quality improvement initiative of a Canadian university and a regional health authority and became a non-profit clinical program in 2021 (Dames et al., 2025). Its main program embeds three intramuscular or sublingual ketamine sessions in a 12-week community of practice, a structured therapy group that meets weekly for preparation, learning and integration. A multidisciplinary team delivers the program, which its protocol describes as weaving Western clinical frameworks with Indigenous knowledge systems, with Indigenous Elders taking part where appropriate. The protocol describes ketamine as an amplifier of change and gives the ceremonial framing, the group process and the curriculum the central role (Dames et al., 2025). As of October 2026, its website described it as a non-profit health care practice and registered charity and listed group and individual programs with psychedelic-assisted therapy sessions, a return-to-work program and clinician training in ketamine-assisted therapy; it did not list a psilocybin program (Roots to Thrive, 2026).
7.1.2 What Its Evaluations Studied
Table 10.4 lists published evaluations of its programs found in October 2026. Its outcome evaluations of ketamine-assisted therapy concern health care workers and firefighters with post-traumatic stress, depression or anxiety; a 2023 qualitative report describes a pilot program for Indigenous participants run in partnership with the Snuneymuxw First Nation; and a 2026 qualitative report describes psilocybin groups for people with terminal illness. None of these evaluations tests treatment for substance use disorders.
Table 10.4. Selected published evaluations of Roots to Thrive programs as of October 2026, with their participants, design and main findings.
| Publication | Participants | Design | Main findings |
|---|---|---|---|
| Dames, Kryskow & Watler, 2022 | 94 health care providers with PTSD, depression or anxiety, three cohorts | Quality improvement cohort, before and after | Anxiety improved in 91%, depression in 79%; 86% of those screening positive for PTSD screened negative at the end of the program |
| Tsang et al., 2023 (chart review) | 128 participants, 448 ketamine sessions, September 2020 to December 2021 | Retrospective safety review | No dropouts; blood pressure raised after 49.2% of sessions; nausea in 12.1%; effects transient |
| Tsang et al., 2023 (pilot comparison) | 57 health care providers in one group program; 38 chose to add three ketamine sessions | Retrospective non-randomized comparison | All improved; reductions larger with ketamine but not statistically significant |
| Manson et al., 2023 | 8 Indigenous participants and 2 Elders, in a pilot program run with the Snuneymuxw First Nation | Qualitative study | Participants attributed benefits to the program and stressed Indigenous team members and traditional approaches to healing |
| Dames et al., 2025 | More than 750 participants to date | Clinical protocol | Describes the model; states it has not been tested in randomized trials |
| Tsang et al., 2025 | Firefighters with treatment-resistant mental health conditions | Prospective observational case series | Anxiety and PTSD scores fell after the 12-week program, and the reductions persisted six months later |
| Tsang et al., 2026 | 25 people with terminal illness, 31 psilocybin sessions, November 2021 to November 2022 | Qualitative study | Participants described greater trust in self and an improved outlook; access through section 56 exemptions and the Special Access Program |
The pilot comparison deserves attention. All 57 health care providers took the group program together, and the 38 who chose to add ketamine showed larger reductions in depression, anxiety and post-traumatic stress scores than the 19 who took the group program alone, and the differences were statistically nonsignificant; the authors took the improvement among all participants to suggest that the benefit came from the group program, with or without ketamine (Tsang et al., 2023). An earlier conference abstract of the pilot, with 49 practitioners (14 without and 35 with ketamine), also found no significant difference between the groups (Tsang et al., 2023). These results fit the protocol’s view that the group does much of the work and leave open whether ketamine adds benefit. Participants chose whether to receive ketamine, so the comparison cannot settle the question.
7.2 Judging Claims About Community Models
Interactive 10.2 asks you to judge which claims about these community models the evidence supports.
For each claim, choose the description that best fits the evidence presented in this section and in Lesson 8, then check your answers.
The pattern across the interactive is that community models have generated encouraging uncontrolled evidence in populations other than people with substance use disorders, and the non-randomized comparison within Roots to Thrive found no statistically significant added benefit from ketamine. That conclusion supports further study and gives no basis for describing these models as treatments for addiction. Research in this area is evolving, and well-designed studies led with the communities involved could change this picture.
A hypothetical promotional sentence reads: “Group ketamine therapy, proven in British Columbia, now offers a new path out of addiction.” Using Table 10.4, rewrite it in two accurate sentences that name the populations studied and the designs used, and state whether any of these evaluations tested treatment for substance use disorders.
Summary and Transition
This section showed that ayahuasca combines DMT with monoamine oxidase inhibitors, that both DMT and harmaline were Schedule III substances in Canada as of October 2026, and that its use runs through Indigenous traditions, Brazilian churches, retreats and research. Studies of people who take ayahuasca in religious rituals found fewer alcohol problems than in comparison groups, two systematic reviews found mainly observational evidence, which they judged promising, and called for controlled trials, and the placebo-controlled trials described here concerned depression and social anxiety. The 2013 study in a First Nations community reported improvements among 12 participants in hope, empowerment, mindfulness and some substance use, with the limits of an uncontrolled design. Religious exemptions protect worship and do not establish therapeutic effect. Roots to Thrive’s group model has uncontrolled evaluations among health care workers, firefighters, Indigenous participants and people with terminal illness, and a non-randomized comparison found no statistically significant added benefit from ketamine. Section 4 brings the four kinds of evidence together and asks how psychedelic approaches would sit beside standard treatment.
Key Takeaways
- 6.1 Ayahuasca combines DMT from the leaves of Psychotria viridis with harmine, harmaline and tetrahydroharmine from the vine Banisteriopsis caapi, which inhibit monoamine oxidase A so that oral DMT can reach the brain. It is used in Amazonian Indigenous traditions, Brazilian churches, retreats and research, and in Canada, as of October 2026, both DMT and harmaline were Schedule III substances, so possessing ayahuasca without an exemption or other authorization was prohibited.
- 6.2 Studies of people who take ayahuasca in religious rituals found lower alcohol use and psychiatric scores than in comparison groups, and a 2016 systematic review found reductions in dependence symptoms or substance use in four of five observational human studies and called for controlled studies. These comparisons cannot separate ayahuasca from membership in a religious community, and the placebo-controlled trials of ayahuasca described here concerned depression and social anxiety, not substance use.
- 6.3 The 2013 study of the Working with Addiction and Stress retreats in a rural First Nations community in British Columbia reported improvements among 12 participants in hope, empowerment, mindfulness and problematic cocaine use six months after four days of group counselling and two ceremonies. With no comparison group, self-reported use and a combined program, no single element can be credited, and Indigenous-led ethical principles bear on who designs such research and how its results are shared.
- 6.4 In the United States, the 2006 Supreme Court decision in Gonzales v. O Centro Espírita Beneficente União do Vegetal opened the way for the church’s sacramental use of hoasca, and in Canada section 56 exemptions have allowed ayahuasca churches to practise since 2017. A religious exemption protects worship, which is a different activity from offering a health treatment, and it is not evidence that ayahuasca treats addiction.
- 7.1 Roots to Thrive in Nanaimo embeds three ketamine sessions in a 12-week community of practice whose group process, curriculum and ceremonial framing play the central role. Its published evaluations concern health care workers and firefighters with post-traumatic stress, depression or anxiety, Indigenous participants and people with terminal illness, using before-and-after, retrospective, non-randomized and qualitative designs, and none tests treatment for substance use disorders.
- 7.2 A claim about a community model should be matched to the evidence behind it: a randomized comparison, uncontrolled evidence only, or no evidence. In the non-randomized comparison within Roots to Thrive, those who chose to add ketamine showed larger reductions than those who did not, but the differences were not statistically significant, so these models warrant further study but cannot yet be described as treatments for addiction.
Reflection
A hypothetical regional health planner in British Columbia says that “community-based psychedelic care has been shown to work for addiction,” citing two examples. First, Thomas et al. (2013) reported on 12 participants in “Working with Addiction and Stress” retreats delivered in a rural First Nations community, which combined four days of group counselling with two expert-led ayahuasca ceremonies; at six months, hopefulness, empowerment and mindfulness had improved, and problematic cocaine use had fallen, while cannabis and opiate use had not. Second, Roots to Thrive in Nanaimo reported improvements among 94 health care providers with post-traumatic stress, depression or anxiety after a 12-week group program with three ketamine sessions, and a non-randomized comparison of 57 health care providers found that reductions with ketamine were larger than with the group program alone but not statistically significant. In 150 to 200 words, respond to the planner, saying what each example shows, what it does not show, and one principle that should guide any expansion of programs drawing on Indigenous ceremony; you may draw on the eight principles that an Indigenous-led group proposed for Western psychedelic research (Reverence, Respect, Responsibility, Relevance, Regulation, Reparation, Restoration and Reconciliation; Celidwen et al., 2023).
Minimum 20 characters required.
Question 1: Why is DMT active when people drink ayahuasca?
Question 2: Thomas et al. (2013) studied the “Working with Addiction and Stress” retreats in a rural First Nations community in British Columbia. Which statement matches the study as its authors reported it?
Question 3: What did the United States Supreme Court decide in the 2006 hoasca case brought by the União do Vegetal?
Question 4: Which statement about Roots to Thrive’s published evaluations is accurate?
What the Evidence Shows and How Psychedelic Approaches Would Sit Beside Standard Treatment
Content note
This section discusses deaths among people with opioid use disorder, including deaths from overdose. If you would like support, the 9-8-8 Suicide Crisis Helpline can be reached by call or text at 9-8-8, and the Hope for Wellness Help Line for Indigenous people is available at 1-855-242-3310.
Introduction
Sections 1 to 3 examined four bodies of evidence: classic psychedelics and ketamine for alcohol and tobacco use, ibogaine for opioid use disorder, ayahuasca, and group models of care. This section steps back to ask three questions. What does the evidence, taken together, show? What does it not yet show? And if psychedelic approaches prove useful, how would they fit alongside the treatments that people with substance use disorders already receive? HSCI 402 develops those standard treatments and harm reduction services; this section uses them as benchmarks and points of connection.
Learning Objectives
- 8.1 Summarize the strength of the evidence for each psychedelic approach to addiction covered in this lesson, as of October 2026.
- 8.2 Identify the main threats to the validity of psychedelic addiction trials and give an example of each from the studies in this lesson.
- 9.1 Compare psychedelic trial results with benchmarks from standard treatments for alcohol, tobacco and opioid use disorders, and explain why such comparisons need care.
- 9.2 Convert an odds ratio into an expected proportion, a risk difference and a number needed to treat at a given baseline risk.
- 9.3 Describe where psychedelic-assisted therapy could sit beside standard treatment and harm reduction in Canada, and the conditions it would need to meet.
8.1 What the Evidence Shows
The evidence differs sharply across substances and indications. One way to summarize it is to ask, for each approach, how far up the ladder of research designs it has climbed.
8.1.1 The Evidence Ladder
Figure 10.4 places each approach on a ladder from uncontrolled studies to regulatory approval, using the studies this lesson describes. As of October 2026, psilocybin and ketamine in alcohol use disorder had reached phase 2 randomized trials, and phase 3 trials of ketamine with psychological therapy (ISRCTN85955128; ISRCTN registry, 2026) and of psilocybin (NCT07638553; ClinicalTrials.gov, n.d.) in alcohol use disorder were recruiting. The LSD trials of the 1960s were pooled in a meta-analysis of six small randomized trials, psilocybin for smoking had an unblinded pilot randomized trial, ibogaine and ayahuasca rested mainly on observational studies, and the group ketamine model had been evaluated in other populations. A 2025 overview of systematic reviews called the evidence for psychedelics in substance use disorders preliminary, noted a high risk of bias, and judged their use for these disorders experimental (Correa da Costa et al., 2025). With trials under way, these positions may change quickly.
8.1.2 What a 2026 Meta-Analysis Found
A meta-analysis published in October 2026 pooled six modern trials of psilocybin-assisted therapy with 202 participants, three of them randomized (Florineth et al., 2026). Drinking fell substantially within groups and stayed lower over follow-up of 4 to 28 weeks. Against control conditions, psilocybin showed moderate short-term advantages on drinks per day, the percentage of drinking days and the percentage of heavy drinking days, but at longer follow-up only the percentage of drinking days remained significantly better. People given psilocybin had higher odds of maintaining abstinence (odds ratio 2.78), and adverse events were mostly mild and transient. The authors described the evidence as mixed and the signals of efficacy as preliminary.
Interactive 10.3 lets you browse the studies behind Figure 10.4. Filter by substance or by design to see how many randomized comparisons support each approach.
Of the 16 studies in the explorer, eight are randomized. The two largest modern trials randomized 95 and 96 participants, the other modern randomized trials 82 or fewer, and the LSD meta-analysis pooled 536 participants from six trials of the 1960s. The other eight studies are observational or uncontrolled, and most of the randomized trials are pilots or phase 2 trials designed to estimate effects for larger trials.
8.2 What the Evidence Does Not Yet Show
Each trial in this lesson reported its own limits, and several recur; a 2023 review described similar problems across psychedelic research (van Elk & Fried, 2023). The accordion below names eight threats to validity, each with an example from this lesson; Lesson 5 explained functional unblinding and expectancy in detail.
In the 2022 psilocybin trial, participants guessed their allocation correctly after 93.6 percent of first sessions (Bogenschutz et al., 2022), and in the French feasibility trial four control participants declined their second session after guessing (Luquiens et al., 2025). The smoking trial was unblinded by design (Johnson et al., 2026). A 2026 systematic review found functional unblinding pervasive among participants and raters in psychedelic randomized trials (Orsini et al., 2026). Expectancy could account for part of each result.
The smoking trial’s odds ratio of 6.12 rested on 4 successes in the patch group, and its confidence interval ran from about 2 to 23 (Johnson et al., 2026). A few more successes in the control group would have changed the estimate considerably.
Trials excluded people with major psychiatric disorders and other drug use disorders, the smoking trial enrolled psychiatrically healthy adults, and 78.9 percent of participants in the 2022 alcohol trial were non-Hispanic White (Bogenschutz et al., 2022). Many people seen in addiction services in Canada would not have been eligible.
Drinking outcomes rested mainly on the timeline followback, and a biological check was available for only 53.8 percent of participants in the 2022 psilocybin trial (Bogenschutz et al., 2022). The ibogaine and ayahuasca studies relied on self-report entirely or almost entirely.
Every trial delivered the drug inside a program of psychotherapy, and control groups improved substantially, with heavy drinking days falling by more than half in the control group of the 2022 psilocybin trial (Bogenschutz et al., 2022). Of the trials in this lesson, only KARE varied the therapy and the drug separately (Grabski et al., 2022).
The smoking trial was funded by the Heffter Research Institute, the Beckley Foundation and a private donor, and its lead investigator reported grants and fees from companies developing psychedelics for work outside the trial (Johnson et al., 2026). Such ties do not invalidate a trial, and they make independent replication and full reporting of negative trials, such as the Swiss trial (Rieser et al., 2025), more important.
Trials of a few dozen participants cannot detect rare serious harms, as Calculator 10.3 showed. For ibogaine, the risk of fatal arrhythmia dominates any judgement of benefit (Section 2).
9.1 Benchmarks from Standard Treatment
A new treatment is judged against what already exists. For each of the three substances in this lesson, Canadian guidelines and systematic reviews describe effective standard treatments, and Table 10.5 summarizes benchmark findings.
Table 10.5. Benchmarks from standard treatments for alcohol, tobacco and opioid use disorders, with the source of each.
| Condition | Standard treatment | Benchmark finding | Source |
|---|---|---|---|
| Alcohol use disorder | Naltrexone or acamprosate as first-line medications, with psychosocial treatment | NNT to prevent return to any drinking: 12 for acamprosate and 20 for naltrexone; to prevent return to heavy drinking: 12 for naltrexone | Jonas et al., 2014; Wood et al., 2023 |
| Tobacco use disorder | Nicotine replacement, varenicline, bupropion and counselling | Standalone nicotine replacement: about 8% prolonged abstinence at six months in a recent meta-analysis | As reported by Johnson et al., 2026 |
| Opioid use disorder | Opioid agonist treatment: buprenorphine-naloxone first line, methadone as alternative | All-cause mortality 11.3 per 1,000 person-years during methadone treatment and 36.1 after leaving it; 4.3 and 9.5 for buprenorphine, from three cohorts (rate difference 5.2, 95% CI −1.0 to 11.4); all unadjusted pooled rates | Bruneau et al., 2018; Sordo et al., 2017 |
The table suggests two lessons. First, a psychedelic approach to opioid use disorder would have to sit beside agonist treatment, because leaving methadone treatment was associated with a roughly threefold higher death rate, although these pooled comparisons are unadjusted and the smaller buprenorphine difference, from three cohorts, has an interval that includes zero (Sordo et al., 2017). Second, the numbers needed to treat in the psilocybin trials look smaller than those for naltrexone and acamprosate: Bogenschutz and colleagues (2022) reported numbers needed to treat of 4.0 to 8.2 for their yes-or-no outcomes. Comparing those figures directly would be misleading, because the outcomes, comparators, follow-up periods and participants differ, and the psilocybin figures come from one trial while the medication figures pool dozens. An odds ratio also translates into different numbers needed to treat depending on how often the outcome occurs without treatment.
9.2 From an Odds Ratio to a Number Needed to Treat
Worked Example 10.4 shows that conversion for the LSD meta-analysis described under 1.1 The LSD Trials and Their Meta-Analysis, using the formula below.
Worked Example 10.4: The LSD odds ratio at two baseline risks
Krebs and Johansen (2012) reported a pooled odds ratio of 1.96, and 38 percent of control patients improved. The expected proportion improving with LSD is (1.96 × 0.38) ÷ (1 − 0.38 + 1.96 × 0.38) = 0.745 ÷ 1.365 = 0.546. The risk difference is 0.546 − 0.38 = 0.166, and the number needed to treat is 1 ÷ 0.166 = 6.0, matching the published figure of six.
If only 10 percent of control patients improved, the same odds ratio would give (1.96 × 0.10) ÷ (1 − 0.10 + 0.196) = 0.179, a risk difference of 0.079 and a number needed to treat of 12.7. The odds ratio is unchanged, and the benefit to a population is about half as large.
Use Calculator 10.4 to repeat the worked example and to apply published odds ratios to other baseline risks.
The hypothetical preset applies the smoking trial’s odds ratio to a baseline of 30 percent and produces a number needed to treat of 2.4. Nothing in the trial shows that the odds ratio would hold in a population with a different baseline, so the preset illustrates the arithmetic and makes no prediction.
A hypothetical colleague argues that psilocybin is “three times better than naltrexone” because Bogenschutz et al. (2022) reported a number needed to treat of 4.3 for abstinence in the final month, while Jonas et al. (2014) reported a number needed to treat of 12 for naltrexone and return to heavy drinking. Name three differences between the two figures that make the comparison misleading, using 8.2 What the Evidence Does Not Yet Show and Table 10.5.
9.3 Where Psychedelic Approaches Would Sit
If larger trials confirm a benefit, psychedelic-assisted therapy would join a system of care that already includes harm reduction services, psychosocial treatment, medications and long-term recovery supports. Figure 10.5 shows one way it could fit, consistent with the evidence in this lesson and the course’s view that the therapy is a phase-based process embedded in a relationship.
Figure 10.5 implies several conditions. The first is evidence: phase 3 or equivalent trials with active comparators or credible blinding, participants who resemble those seen in services, and follow-up of at least a year. The second is a lawful route. In Canada, psilocybin and MDMA have no market authorization for any indication as of October 2026, and a practitioner may request them through the Special Access Program only for a serious or life-threatening condition; Health Canada’s February 2023 notice states that such requests are assessed case by case, that conventional treatments must have failed, be unsuitable or be unavailable, and that the program is not a route for research (Health Canada, 2023). Lesson 7 explained the legal limits on offering preparation or integration for unauthorized use, and Lesson 11 develops access routes.
The third condition is safety. Screening for cardiac and psychiatric risk (Lesson 6), review of medications such as antidepressants and other drugs that interact with psychedelics (Lesson 3), and medical monitoring would apply to every program, and for opioid use disorder the program would need to protect against overdose after lost tolerance (see 5.1 Withdrawal Management Alone and the Risk of Overdose). The fourth is integration with ongoing care: every trial in this lesson embedded the drug in psychotherapy, so a psychedelic session would be one phase of a longer program, followed by integration and a connection back to the person’s regular treatment (Lesson 7). The fifth is equity. Psychedelic-assisted therapy requires hours of skilled staff time per person and trial samples have been narrow, so a program would need to show that it reaches the people who carry the heaviest burden of substance use harms, a question that HSCI 402 and Lesson 12 develop.
The course’s practice standard favours a regulated public health model, with licensed providers, training standards and public funding where evidence supports it, over commercial provision without standards. That is one side of a policy debate. Others argue that strict standards set early can slow access, that private investment pays for large trials such as the company-run phase 3 trials of psilocybin in depression described in Lesson 8, and that group models such as Roots to Thrive are designed to be scalable (Dames et al., 2025). Lessons 11 and 12 present those arguments in full. The box below gathers questions that apply whichever model prevails.
Questions to ask of any psychedelic addiction program
Ask which substance use disorder the evidence concerns and whether the program’s clients resemble the trial participants; whether the evidence includes a randomized comparison and how the blind was protected; how long benefits lasted and what happened to people who returned to use; what screening, monitoring and emergency care the program provides; how the program connects clients to standard treatment, including agonist treatment for opioid use disorder; by what legal route the drug is provided; and who can afford it.
A policy analyst in British Columbia asks for a briefing of about 300 words on three proposals: (a) public funding for British Columbia sites in phase 3 trials of psilocybin-assisted therapy for alcohol use disorder; (b) offering group ketamine-assisted therapy modelled on Roots to Thrive to people with alcohol use disorder; and (c) reimbursing ibogaine treatment at clinics outside Canada for people with opioid use disorder. For each proposal, use the whole lesson to (1) state the strongest evidence that bears on it and the design of that evidence (Sections 1 to 3 and Figure 10.4); (2) name the main safety or validity concern (Section 2 and 8.2 What the Evidence Does Not Yet Show); (3) describe how it would sit beside standard treatment, including agonist treatment and harm reduction (9.3 Where Psychedelic Approaches Would Sit and 5.1 Withdrawal Management Alone and the Risk of Overdose); and (4) recommend whether the province should fund it now, fund research on it, or decline, with one sentence of justification.
Summary
This section brought the lesson’s evidence together. As of October 2026, the strongest evidence described here consisted of phase 2 trials of psilocybin and ketamine in alcohol use disorder, with phase 3 trials recruiting, a meta-analysis of 1960s LSD trials and a pilot trial in smoking, while ibogaine and ayahuasca rested mainly on observational studies and published evaluations of group ketamine models concerned other populations; a 2025 overview of systematic reviews called the evidence preliminary. Recurring threats to validity include broken blinding, small samples, selected participants, self-reported outcomes, short follow-up, the therapy confound and ties to funders. Standard treatments provide benchmarks and, for opioid use disorder, a reduction in mortality that any new approach would need to protect. If psychedelic-assisted therapy earns a place, it would sit inside psychotherapy and alongside standard care, reached through lawful routes, with safety and equity built in. The final assessment that follows asks you to apply these judgements.
Key Takeaways
- 8.1 As of October 2026, the strongest evidence described here came from phase 2 trials of psilocybin and ketamine in alcohol use disorder, with phase 3 trials recruiting. The highest rungs reached were phase 2 trials of psilocybin and ketamine in alcohol use disorder, a meta-analysis of 1960s LSD trials and an unblinded pilot trial in smoking, while ibogaine and ayahuasca rested mainly on observational studies. A 2026 meta-analysis described the psilocybin evidence in alcohol use disorder as mixed and its signals of efficacy as preliminary.
- 8.2 Recurring threats to validity include broken blinding and expectancy, small samples with fragile estimates, selected participants, self-reported outcomes, limited durability, the therapy confound, funding and allegiance, and rare harms that small trials cannot detect. Each has an example in this lesson, such as allocation guessed after 93.6 percent of first sessions in a 2022 psilocybin trial and an odds ratio of 6.12 resting on four successes in a control group.
- 9.1 Standard treatments set the benchmarks: naltrexone or acamprosate for alcohol use disorder, nicotine replacement and other aids for tobacco, and opioid agonist treatment, where leaving methadone treatment was associated with a roughly threefold higher death rate in unadjusted pooled data. The psilocybin trial’s numbers needed to treat look smaller than those for naltrexone and acamprosate, but outcomes, comparators, follow-up periods and participants differ, so direct comparison misleads.
- 9.2 An odds ratio converts to an expected proportion through p1 = (OR × p0) ÷ (1 − p0 + OR × p0), and the number needed to treat is one divided by p1 − p0. The LSD odds ratio of 1.96 gives a number needed to treat of 6.0 when 38 percent of controls improve and 12.7 when 10 percent do, so the same odds ratio implies different benefits at different baseline risks.
- 9.3 If larger trials confirm a benefit, psychedelic-assisted therapy would sit inside a program of preparation, a supervised session and integration, added to standard treatment and harm reduction and never replacing agonist treatment for opioid use disorder. It would need phase 3 evidence, a lawful route such as clinical trials or the Special Access Program, safety screening and protection against overdose after lost tolerance, integration with ongoing care, and equitable access.
Reflection
A hypothetical private clinic in a Canadian city (not a real service) advertises “ketamine therapy for alcohol addiction, proven in trials,” offering one infusion with no preparation sessions, no follow-up and no contact with the client’s physician, at a cost of $900 paid by the client. The clinic cites the KARE trial, in which 96 people with severe alcohol use disorder received three weekly ketamine or saline infusions with either relapse prevention therapy or education, and ketamine produced 10.1 percentage points more days abstinent at six months with no difference in relapse. Using the questions to ask of any psychedelic addiction program under 9.3 Where Psychedelic Approaches Would Sit (which disorder the evidence concerns, whether it includes a randomized comparison, how long benefits lasted, what screening and monitoring are provided, how the program connects to standard treatment, what legal route is used and who can afford it), write 150 to 200 words assessing the clinic’s claim and listing two changes that would bring its program closer to the evidence.
Minimum 20 characters required.
Question 1: As of October 2026, which approach to addiction in this lesson had reached the highest level of controlled evidence for a substance use disorder?
Question 2: An odds ratio of 1.96 is applied to a hypothetical control group in which 10 percent improve. What is the approximate number needed to treat?
Question 3: Why would directly comparing an NNT of 4.3 from Bogenschutz et al. (2022) with an NNT of 12 for naltrexone from Jonas et al. (2014) be misleading?
Question 4: Which description best fits how psychedelic-assisted therapy would sit beside treatment for opioid use disorder, based on this lesson?
Final Assessment
Bringing It All Together
This lesson followed psychedelic approaches to addiction from the LSD trials of the 1960s to the trials, programs and policies of 2026. For alcohol use disorder, a meta-analysis of six LSD trials found a consistent short-term benefit that faded by 12 months, a 2022 psilocybin trial reduced heavy drinking days by about 14 percentage points against an active placebo, a Swiss trial of a single dose found no difference, and two ketamine trials found more abstinence. For smoking, an unblinded pilot trial found that about four times as many people quit with psilocybin as with nicotine patch. Mechanisms rest mainly on correlations with the acute experience and on animal research, and in people they remain largely speculative.
For opioid use disorder, ibogaine rests mainly on uncontrolled cohorts while its blockade of cardiac hERG channels and a series of deaths make safety the central question; small studies without events still leave wide uncertainty, and Canadian guidance against withdrawal management alone applies to single-session treatment. Reviews describe ayahuasca research as mostly observational, the 2013 study in a First Nations community reported encouraging changes among 12 participants, and religious exemptions protect worship without establishing therapeutic effect. Roots to Thrive’s group model has been evaluated among health care workers, firefighters and other groups, and none of the published evaluations described here tests treatment for addiction.
Taken together, as of October 2026 the strongest evidence described here came from phase 2 trials, with phase 3 trials of ketamine and psilocybin in alcohol use disorder recruiting, and a 2025 overview of systematic reviews called the evidence preliminary and the use of psychedelics for substance use disorders experimental. If psychedelic-assisted therapy earns a place in addiction care, the evidence points to an addition within psychotherapy, alongside standard medications and agonist treatment, reached through lawful routes and designed for safety and equity.
Key Takeaways from this lesson
- Krebs and Johansen (2012) pooled six LSD trials from 1966 to 1970 with 536 participants and found an odds ratio for improvement of 1.96 at first follow-up that was no longer significant at 12 months. Bogenschutz et al. (2022) found 9.7 versus 23.6 percent heavy drinking days with psilocybin and an active placebo, while participants guessed their allocation after 93.6 percent of first sessions, and a Swiss phase 2 trial of a single psilocybin dose found no difference from placebo, and a 2026 meta-analysis of six trials described the evidence for psilocybin in alcohol use disorder as mixed and preliminary. Two randomized ketamine trials found more abstinence in alcohol use disorder, and the 2026 smoking trial was an unblinded pilot of 82 adults in which 40.5 percent with psilocybin and 10.0 percent with nicotine patch showed prolonged abstinence.
- Published trial data can be turned into differences in heavy drinking days, odds ratios, risk differences and numbers needed to treat. In the 2026 smoking trial, a risk ratio of about 4 and a number needed to treat of 3.3 sat beside an odds ratio of about 6, because odds exaggerate a ratio when the outcome is common.
- Proposed mechanisms of change rest mainly on correlations with the acute experience in small samples and on animal research, and a secondary analysis found that ketamine’s psychoactive effects did not mediate its benefit.
- Evidence for ibogaine in opioid use disorder comes mostly from uncontrolled studies, as a 2023 systematic review found, while ibogaine and noribogaine block hERG channels, prolong the QT interval and have been linked to deaths. A study with no serious events among n participants leaves an upper 95 percent limit of about 3 divided by n, so small studies cannot establish safety against rare fatal events.
- Canadian guidance advises against withdrawal management alone for opioid use disorder because lost tolerance raises the risk of overdose death, which bears directly on single-session ibogaine treatment. Texas and the United States government have committed public funds to drug development to generate the controlled evidence that is missing.
- Ayahuasca combines DMT with monoamine oxidase inhibitors, both DMT and harmaline were Schedule III substances in Canada as of October 2026, and reviews describe its evidence in addiction as observational, including the 2013 study among 12 participants in a First Nations community. Religious exemptions protect worship without establishing therapeutic effect.
- Roots to Thrive’s published evaluations concern health care workers, firefighters and other groups, none of them tests treatment for substance use disorders, and a non-randomized comparison found no statistically significant added benefit from ketamine.
- As of October 2026, the strongest evidence described here came from phase 2 trials, and recurring threats to validity, including broken blinding, small and selected samples, self-reported outcomes, short follow-up and the therapy confound, limit confidence in the trials that exist.
- Benchmarks from standard treatment cannot be compared directly with psychedelic trial results, because an odds ratio implies different numbers needed to treat at different baseline risks and the outcomes and participants differ. If psychedelic-assisted therapy earns a place, it would be an addition within psychotherapy, alongside standard medications and agonist treatment, reached through lawful routes and designed for safety and equity.
Core Concepts Reviewed
Section 1: the LSD meta-analysis, percentage of heavy drinking days, the timeline followback, active placebos, prolonged and point prevalence abstinence, odds ratios, risk ratios, risk differences, numbers needed to treat, pilot trials, ketamine trials and proposed mechanisms of change.
Section 2: ibogaine and noribogaine, Bwiti practice, observational cohorts, regression to the mean, the QT interval, hERG channels, torsades de pointes, the case series of deaths, the rule of three, magnesium co-administration, withdrawal management alone and state-funded drug development.
Section 3: ayahuasca’s composition and monoamine oxidase inhibition, its Canadian legal status, studies of ritual participants, the 2013 retreat study in a First Nations community, ethical principles of traditional Indigenous medicine, religious exemptions and Roots to Thrive’s community of practice model.
Section 4: the evidence ladder, threats to validity, benchmarks from standard treatment for alcohol, tobacco and opioid use disorders, converting odds ratios to numbers needed to treat, access routes in Canada and conditions for a place in care.
The final reflection asks you to recommend one research project using evidence from all four sections.
Reflection
Imagine that a federal research funder can support one of three hypothetical projects for five years. Project A is a multi-site phase 3 trial of psilocybin-assisted therapy for alcohol use disorder, building on a phase 2 trial that reduced heavy drinking days from 23.6 to 9.7 percent and a later trial that found no difference from placebo. Project B is a randomized trial of magnesium-ibogaine for opioid use disorder, building on uncontrolled cohorts of 14 to 191 people, a case series of 19 deaths and an open-label study of 30 veterans without serious events. Project C is a randomized trial of a group-based ketamine-assisted program for alcohol use disorder in British Columbia, building on uncontrolled evaluations among health care workers and a non-randomized comparison that found no statistically significant added benefit from ketamine. In 200 to 250 words, recommend one project, using evidence from all four sections of the lesson: the strength of existing evidence, the main safety concern, how the treatment would sit beside standard care, and one equity consideration.
Minimum 30 characters required.
Final Knowledge Assessment
Question 1: Which pair of findings best illustrates conflicting results among psilocybin trials in alcohol use disorder?
Question 2: Which feature, shared by nearly every positive trial in this lesson, complicates attributing benefit to the drug alone?
Question 3: A hypothetical clinic advertises that “one ibogaine session ends opioid addiction.” Which combination of evidence from this lesson most directly challenges the claim?
Question 4: Which statement correctly contrasts the odds ratio and the risk ratio in the 2026 smoking trial (17 of 42 versus 4 of 40)?
Question 5: Why does Calculator 10.3 matter for judging ibogaine’s safety?
Question 6: As of October 2026, which statement about ayahuasca in Canada is correct?
Question 7: Which conclusion about mechanisms of change is best supported by the evidence in Section 1?
Question 8: What does the 2026 meta-analysis of psilocybin-assisted therapy for alcohol use disorder (Florineth et al.) add to the picture?
Question 9: Which group was studied in Roots to Thrive’s 2022 cohort report on group ketamine-assisted therapy?
Question 10: In a psychedelic trial, participants guessed their allocation correctly after more than 90 percent of sessions. Which threat to validity does this most directly raise?
Question 11: Which benchmark from standard care is most relevant when judging any new approach to opioid use disorder?
Question 12: Which route could, as of October 2026, allow a Canadian practitioner to request psilocybin for a patient outside a clinical trial?
Question 13: Which statement best describes the contribution and limits of Thomas et al. (2013)?
Question 14: Why is a religious exemption for an ayahuasca church weak evidence about treatment for addiction?
Question 15: Which policy development funds research on ibogaine without endorsing current ibogaine practice?
Glossary: Key Terms, People & Frameworks
📚 Reference page, available throughout the lesson
Terms, frameworks and people introduced in this lesson on psychedelic approaches to addiction, with short definitions.