Ready for Discharge Is Not Ready for Practice

November 3, 2026 | The Psychiatric Record | Treatment Evidence

On October 26, JAMA published a randomized trial of a single nasal dose of mebufotenin benzoate, the compound better known as 5-MeO-DMT, in adults with treatment-resistant depression. It found a meaningful four-week benefit over a very low dose, and most participants were judged ready for discharge within two hours. Under the trial's own rules, none could leave before four.

The phase 2b trial randomized adults aged 18 to 75 with moderate to severe major depressive disorder that had not responded to 2 to 5 antidepressant trials in the current episode, at 38 sites in Australia, Europe, and the US. To test the drug as monotherapy, participants taking antidepressants completed a supervised washout at least two weeks before dosing. Then 193 received one dose of mebufotenin, 8 mg (46), 12 mg (73), or 0.3 mg (74), with manualized psychological support before, during, and after.

The primary end point was change on the Montgomery-Åsberg Depression Rating Scale (MADRS, 0 to 60) at day 29, 12 mg against 0.3 mg. Scores fell by 11.1 points with 12 mg and 5.5 with 0.3 mg; the least-squares mean difference from a repeated-measures model was -5.3 points (95% CI, -8.9 to -1.8). The 8 mg dose, a secondary comparison, fell 12.5 points. Day 29 response rates were 35% (8 mg), 34% (12 mg), and 12% (0.3 mg); remission rates were 26%, 18%, and 8%. Both higher doses stayed ahead of the comparator through day 57, the last blinded visit. Using the rounded 12 mg and 0.3 mg response rates, our calculated number needed to treat is roughly 5.

Readiness was checked every 30 minutes from 90 minutes after dosing, using a nine-item checklist in the paper's supplement: the participant is fully responsive and oriented, acute psychedelic effects have completely subsided, blood pressure and pulse are normal or only slightly elevated, breathing and temperature are normal, gait is stable, side effects are mild to moderate and need no medical monitoring, there is no acute suicidal ideation or intent, distress has settled enough that the participant feels safe, and the assessor judges discharge safe. A healthcare provider or study clinician verified the final questionnaire. Separately, the Methods state, "The earliest participants could be discharged was 4 hours after dosing."

By two hours, 85% of participants on 8 mg and 78% on 12 mg met readiness. The supplement's timing table reports 6 of 46 and 15 of 73 later than two hours, with maximums of 243 and 261 minutes. Those cumulative counts leave one participant unaccounted for in each active arm, missed assessments carried the previous result forward, and the table reports readiness, not when anyone actually left. We found no distribution of actual departure times.

The authors conclude that mebufotenin "may improve treatment access and convenience, requiring a shorter monitoring period and fewer clinic visits compared with psilocybin and esketamine, respectively." That may prove right, but the trial had no psilocybin or esketamine arm, and the two-hour figures measure different things. For treatment-resistant depression, esketamine's label sets a floor: patients "must be monitored for at least 2 hours at each treatment session." The mebufotenin figure is the share meeting readiness criteria by two hours, under a protocol requiring a stay of at least four hours.

Visit burden is also more than dosing sessions. Esketamine's label calls for twice-weekly sessions in weeks 1 to 4, weekly in weeks 5 to 8, then weekly or every two weeks, individualized to the least frequent schedule that maintains response. The developer's phase 3 extensions plan individualized 8 mg retreatment "at 8-week or 12-week intervals," which could sharply reduce dosing visits. But this trial also scheduled three preparation and three follow-up support visits of about 60 to 90 minutes each, some allowed remotely, plus assessment visits. What a marketed care package would require is not yet known.

Treatment-emergent adverse events occurred in 76% (8 mg), 85% (12 mg), and 73% (0.3 mg), and 99.7% were mild or moderate in severity. In the registry's coded core-period counts, nausea was reported by 29 of 73 on 12 mg and 2 of 74 on 0.3 mg, and panic attacks by 6 and 0. Blood pressure and heart rate rose after dosing: peak mean systolic pressure was 147 and 146 mm Hg in the active groups, and the paper reports return to the reference range within one hour. Those are group means, not individual maximums. The authors reported two serious events in the blinded period, both judged unrelated to treatment. This sample cannot rule out uncommon serious harms.

The Columbia suicide scale was given at every visit. In the blinded period, the supplement reports no suicidal behavior in any group and no suicidal ideation with intent in either active group. That is reassuring for this period; it is not evidence of suicide prevention. Participants who completed the blinded period could enter an open-label extension assigned to a second dose regimen; 107 entered, and there the registry lists two serious events of suicidal ideation and one of binge drinking, with the sponsor judging one ideation event related to the drug.

Dosing was designed as monotherapy, but 12 participants started antidepressants during the study (3, 6, and 3 in the 8 mg, 12 mg, and 0.3 mg groups). The paper does not say when or why, and the trial does not establish how the drug behaves when given alongside an antidepressant.

Not well, for participants and clinicians, and the authors say so. The paper reports "significant degrees of functional unblinding of clinicians and participants," with blinding "more preserved" for the independent MADRS raters. In the supplement's treatment-guess data, 98 of 119 participants given an active dose strongly believed they had received active treatment, and 43 of 74 on 0.3 mg strongly believed they had received the comparator. FDA's July 2026 guidance on psychedelic trials names this as a source of expectation bias.

The authors report no correlation between pretreatment expectancy and day 29 MADRS change. That does not exclude expectancy that arose after participants felt the drug. The fair reading is a statistically significant signal, measured by better-blinded raters, in a trial where many participants correctly guessed their assignment.

The trial was funded by Beckley Psytech, an AtaiBeckley company, and sponsor employees were involved in its design and conduct, data collection, analysis and interpretation, manuscript preparation and approval, and the decision to submit. The authors' limitations include the masking problem, the smaller 8 mg group after randomization was changed to prioritize the primary end point, the required antidepressant washout, limited ethnic diversity, possible underreporting of prior psychedelic use, and that the contribution of psychological support was not evaluated.

Mebufotenin is not available for prescription in the US. 5-MeO-DMT is listed in Schedule I, 21 CFR 1308.11(d)(15), and any marketed product would need FDA approval and DEA rescheduling. AtaiBeckley's filings describe breakthrough therapy designation in October 2025, an end-of-phase 2 meeting in February 2026, and two planned pivotal trials, ReConnection-1 (about 350 patients) and ReConnection-2 (about 230), with 12-week placebo-controlled cores. Our ClinicalTrials.gov searches on November 3 did not locate a phase 3 registration. Eli Lilly, which completed its acquisition of AtaiBeckley in September, now holds the program.

A patient who asks about the "two-hour psychedelic" can be told that it is a Schedule I investigational drug available in the US only in research, that one phase 2b trial found a moderate four-week benefit over a very low dose despite substantial functional unblinding among participants, and that a shorter session is a promising feature phase 3 still has to confirm. No one should stop an antidepressant on their own in anticipation of access.

Early readiness is promising. This trial did not establish a two-hour clinical treatment session.

Source note: Trial methods and results come from the JAMA article, its Supplement 3, and the sponsor's ClinicalTrials.gov results. We did not review the full protocol, statistical analysis plan, or data-sharing supplements, or the accompanying editorial. The supplement's readiness table reports readiness times, not actual departure times, and its counts do not sum to the full active groups; we report them as published. Treatment-guess figures are as shown in the supplement, without a summary index. Coded nausea and panic counts and serious events in the extension come from the registry; core serious-event attribution comes from the paper. Company plans are from SEC filings, and the phase 3 statement reflects bounded registry searches. The number needed to treat is our calculation from rounded response rates, without a confidence interval.

Educational Disclaimer: The Psychiatric Record provides general educational information for psychiatric and mental-health professionals. Content does not constitute medical, legal, regulatory, compliance, billing, or other professional advice; does not establish a standard of care; and is not a substitute for independent professional judgment. Mebufotenin is an investigational Schedule I substance and is not approved for any use. Esketamine requirements cited here come from its current FDA labeling and REMS, which clinicians should confirm before relying on them.