A VA pilot dosed only the veteran and treated the couple. With Oregon outcomes data published and MDMA back at the FDA, the delivery model is the question.
Key Takeaways
- The design shift: An eight-session VA pilot paired brief couples therapy with two MDMA sessions for the veteran alone, testing whether dosing one person can treat two.
- The real-world evidence: Oregon’s first multisite cohort, 346 adults at 24 licensed centers, reported serious adverse reactions in 1.2 percent of sessions and lower depression and anxiety at three months.
- The cost constraint: Sessions that occupy two clinicians for a full day are the central obstacle, and every scaling proposal in the field is an attempt to shrink that number.
- The near-term path: An MDMA application is back before the FDA and psilocybin is in Phase 3, which puts staffing and session design in front of operators before any coverage exists.
Figure 1 of the study protocol reads like a production schedule. Eight therapy sessions. One preparation session focused on the medication. Two full-day dosing sessions. Then an integration session the next morning, both partners present. Teams of two co-therapists deliver the sequence. Only one person in the room receives the drug.
That protocol, published in 2024 and registered as NCT05979844, became the first U.S. trial of MDMA-assisted brief cognitive-behavioral conjoint therapy, a couples treatment for post-traumatic stress disorder. Its results appeared this year in the Journal of Traumatic Stress. The sample was tiny and the design uncontrolled, so the findings settle nothing about efficacy. What the protocol does illustrate is a question the field has moved on to. Pharmacology has absorbed a decade of study. The delivery model around it has not: who is in the room, for how long, under whose supervision and at whose expense. That is the part determining whether any of this reaches ordinary behavioral health organizations.
What the VA Couples Pilot Tested
The pilot, led by Leslie A. Morland and colleagues at the Veterans Affairs San Diego Healthcare System, enrolled eight veterans with PTSD and their intimate partners, 16 people in all. The work was funded by a grant from the Heroic Hearts Project, and the study drug was supplied by Resilient Pharmaceuticals, the company whose application now sits at the FDA. Brief cognitive-behavioral conjoint therapy is an existing manualized treatment in which the couple, rather than the patient alone, works through PTSD psychoeducation, communication skills and behavioral approach exercises. Onto that spine the trial grafted two dosing days for the veteran and dyadic integration sessions drawing on integrative behavioral couples therapy.
The single-partner dosing decision is the consequential one. An earlier pilot of the full conjoint protocol gave the medication to both partners and reported a large effect on PTSD symptoms, with five of six patients in sustained remission six months out. Dosing one person halves the monitoring burden and the medical screening, and leaves the partner clear-headed throughout. The authors add a finding that makes the choice look less like a compromise: seven of the eight partners either did not want MDMA or could not take it because of work. Morland’s team reported a large reduction in clinician-rated PTSD severity, an average 16.1-point drop on the CAPS-5 from a baseline mean of 39.13. Five of the eight veterans met the threshold for clinically significant response, three no longer met criteria for the diagnosis, and two reached remission. Relationship satisfaction improved for veterans and partners alike, by 27.9 and 19.8 points on a standard couples measure. Both PTSD scores and satisfaction scores rebounded somewhat during follow-up, and one partner’s satisfaction deteriorated meaningfully. No serious adverse events were reported, and all eight couples finished the full course. Eight couples cannot support more than a hypothesis, which the authors say plainly, and they note that their rates of diagnosis loss and remission ran below those reported in the Phase 3 MDMA trials, closer to what the same team has seen with conjoint therapy alone. This design, though, is the first to treat the relationship as the unit of care while treating one person as the patient.
The schedule is the part operators should read closely. Therapy was compressed, with the first three sessions delivered in a single four-hour block and the next two in a three-hour block, all virtually, by a therapist sitting in a VA office. Dosing days happened in person and ran six to eight hours, with the couple returning the next morning to process it together. Treatment averaged seven weeks. Each couple had a primary therapist throughout and a co-therapist for preparation, dosing and integration, all of them doctoral-level clinicians trained in both protocols.
Oregon’s First Outcomes Data, and the Group Share
Evidence of a different kind arrived in August. Researchers led by P. Todd Korthuis at Oregon Health and Science University published the first multisite look at outcomes in Oregon’s regulated psilocybin program in JAMA Network Open, following 346 adults who completed a session at 24 of the 26 centers holding active licenses. Enrollment ran from November 2024 through March 2026, with follow-up through that June. Eighty-three facilitators supervised the sessions, and retention stayed above 90 percent at every follow-up. The cohort is about 5 percent of everyone served in that window, out of more than 20,000 since 2023.
Among participants reaching three-month follow-up, the share reporting moderate to severe depression fell from 42.2 percent to 16.5 percent. Moderate to severe anxiety dropped over the same period, from 45.1 percent of participants to 13.2 percent. Facilitators or clients reported serious behavioral adverse reactions in four sessions, 1.2 percent, one requiring medical attention, and no serious medical reactions. Seven of the 312 participants at three months, 2.3 percent, described the experience as harmful. PTSD symptoms followed the same direction, with moderate to severe scores falling from 48.0 percent to 16.8 percent. One finding should interest any state weighing a program: facilitators reported only one of the four serious reactions, which the authors read as evidence that reporting rules limited to the day of service understate what happens. Participants selected themselves and had no control group, and Korthuis noted in an OHSU release that they may have been healthier overall than clinical trial populations.
One line in the results speaks directly to the delivery question. Roughly 18.5 percent of participants, 64 of 346, received psilocybin in group sessions rather than individually. Oregon’s program, whose economics Acuity examined this month, permits that configuration. The study did not compare outcomes between the two formats, and nobody else has published that comparison at scale, but the group share is evidence that supervised models are already experimenting with the staffing ratio. The same data show the surrounding labor: 2.4 preparation sessions on average beforehand, and 1.9 integration sessions for the 90.6 percent who attended any.
The Labor Arithmetic Behind Every Psychedelic Delivery Model
The reason formats matter is straightforward. Testifying before a Senate committee in May, VA Secretary Doug Collins put it bluntly, saying MDMA-assisted therapy requires almost 120 hours per patient with two psychiatrists, according to an account of the hearing. The figure has not been formally validated, and much of that time in practice falls to therapists rather than physicians, but it explains why estimates for a full course run above $12,000.
Set that against a behavioral health workforce that is already short-staffed, a subject Acuity covered in its reporting on clinician shortages and burnout. Two clinicians tied up all day with one patient is a staffing pattern few outpatient organizations can absorb at prevailing rates. Every serious proposal for scaling this work is therefore a proposal about labor: shorter manualized protocols, group sessions instead of individual ones, one dosed patient instead of two, or lower-cost facilitators under clinical supervision. Morland’s protocol takes the first and third routes. Oregon’s group sessions take the second.
Where the Regulatory Track Stands
The clinical pipeline has kept moving through all of this. COMPASS Pathways reported in its most recent quarterly filing that it met the primary endpoint in its Phase 3 COMP006 trial in February and released further results in July. In April the FDA granted the company a rolling review and awarded a Commissioner’s National Priority Voucher for its psilocybin candidate in treatment-resistant depression, a designation meant to compress the review window once a filing is complete. The company told investors in May that it expects to finish the application this quarter, and its chief executive said in September that every module except the clinical one had gone in. The agency separately accepted an investigational new drug application in January for a Phase 2b/3 trial in PTSD.
MDMA’s path has been slower. The FDA declined to approve the original application in August 2024, citing durability of benefit, safety characterization and trial design. The sponsor, now called Resilient Pharmaceuticals after operating as Lykos Therapeutics and before that as the public benefit corporation formed by MAPS, resubmitted its application on August 9, first reported by Psychedelic Alpha and since confirmed by MAPS, which funded the original trials but has no role in the current program. Resilient has not discussed the filing publicly. The package reportedly went back without the additional Phase 3 trial the agency had requested, supported instead by outside research and a new Phase 1 study in 32 volunteers. A resubmission is not an acceptance for review, and acceptance is not approval. Both applications are moving through an administration that has made this a stated priority. Executive Order 14401, signed in April, directs the FDA to prioritize review of these treatments and orders the Attorney General to review rescheduling for any Schedule I product that has finished Phase 3 trials for a serious mental health disorder. That second piece matters practically, because a drug cannot be prescribed while it sits in Schedule I whatever the FDA decides. In July the FDA finalized guidance titled Psychedelic Drugs: Considerations for Clinical Investigations, which addresses the therapist-delivered components of these treatments, a signal that the agency regards the therapy wrapped around the molecule as part of what it is evaluating.
Who Would Pay, and Under What Codes
No standing payment mechanism exists for any of this outside clinical trials. Three Category III CPT codes exist, issued by the American Medical Association in 2023 to cover in-person monitoring and intervention during psychedelic medication therapy, as the Petrie-Flom Center at Harvard Law School has described, and they amount to a placeholder rather than a benefit. Category III codes carry no assigned relative values, and payers are not obliged to recognize them. Esketamine offers the nearest working precedent, billed through separate codes for the drug and for the supervision period at certified sites under a risk evaluation and mitigation strategy, a structure that took years to settle.
The more useful comparison for operators may be the collaborative care model, where a defined protocol, a measurable population and dedicated billing codes eventually produced steady Medicaid growth, or the cost-based rates that make Certified Community Behavioral Health Clinic economics work. Both show that structured care can be financed once the unit of service is legible to a payer, and both took years. Meanwhile the reimbursement gap Acuity described in May, in which evidence-based services are paid inconsistently even after the evidence arrives, is the default condition any psychedelic protocol will enter.
There is a further wrinkle specific to the dyadic model. Couples therapy has never been reimbursed well. A protocol treating two people, only one of whom carries the billable diagnosis, runs into the same categorical problems Acuity documented in its coverage of billing silos in co-occurring care, where the patient’s needs cross boundaries the claim form does not recognize. Morland’s team raises this itself, noting that if the drug is approved, coverage rules will likely confine prescriptions to the person carrying the diagnosis. Their single-dosing design anticipates that constraint rather than fighting it.
What Behavioral Health Operators Can Watch
The federal government has become the most active party here. VA funded its first psychedelic-assisted therapy study since the 1960s in December 2024 and has since built a portfolio of roughly 20 active trials supported by more than $23 million in external funding, including an MDMA trial for veterans with both PTSD and alcohol use disorder that began enrolling in May. In July, the VA and HHS signed a five-year memorandum of understanding that includes training clinicians to administer psychedelic medications should federal approval arrive.
That matters practically. Protocols built inside a health system are shaped around existing staffing, documentation and supervision, which makes them easier to hand to another health system than a model designed for a standalone center. Morland’s couples protocol was written with that portability in mind, describing itself as a test of a scalable approach within VA care.
Three markers will show how fast this moves. Whether the FDA accepts the MDMA resubmission, and on what timeline, decides if these questions turn urgent in 2027. Whether COMPASS finishes its rolling submission this quarter, and how fast the DEA moves on rescheduling after any approval, sets the second clock. And whether group and dyadic formats produce outcomes comparable to individual sessions decides whether the labor arithmetic can be solved at all. Until then, supervised psychedelic care remains available only through clinical trials and a handful of state programs, and behavioral health organizations are watching a treatment model still deciding what shape it will take.
