Evolving Guidelines for the Use of Touch During a Clinical Trial of Group Psilocybin-Assisted Therapy

Back and colleagues situate this Technical Report within a context of recent ethical concerns about inappropriate use of touch in psychedelic-assisted therapy and calls from practitioner bodies for clearer guidance. Earlier incidents of boundary violations prompted attention from professional organisations and researchers to issues of consent, autonomy, and risk mitigation, while some experienced psychedelic practitioners and legacy practices (for example Holotropic Breathwork and historic LSD therapy descriptions) argue that carefully used touch can be supportive. The literature, however, lacks empirical guidance about how to use touch safely in contemporary clinical-trial settings, particularly in group retreat formats. This report describes how the study team developed, implemented, and then revised an initial set of touch practice guidelines during a Phase I/II group psilocybin-assisted therapy trial for people with metastatic cancer. The authors present two case events from their third retreat that prompted iteration of the guidelines, the content of the evolved guidelines, and early experience using the revised guidance in a subsequent retreat. Their stated aim is to document empirical cases and team processes to inform further ethical discussion and research on the role of touch in psychedelic therapy.

The work is embedded in a registered Phase I/II clinical trial of a group model of psilocybin-assisted therapy for cancer-related anxiety in patients with metastatic cancer (ClinicalTrials No: NCT05847686) and received institutional review board approval. The trial uses a retreat-based group format in which participants receive three 90-minute group preparation sessions, one 8-hour group psilocybin session conducted in a large room with participants on mattresses, and three 90-minute group integration sessions. Two preparation sessions are online, the third is in person at a 3-day retreat, and integration includes one in-person session on the retreat plus two online sessions. Each participant also has one individual in-person preparation session and one individual in-person integration session; the psilocybin session is audio‑ and video‑recorded. Back and colleagues initially created a set of prescriptive touch guidelines before the study retreats. Key elements were that touch would be framed as supportive for intense experiences, would be limited to hands, arms and possibly feet, that participants would privately specify touch "preferences" in advance (including an option for no touch), that preferences would be respected and not renegotiated during the psilocybin session, and that any touch beyond those areas would be reserved for preventing serious harm. The facilitator team received training on these guidelines, including partner practice, trauma history considerations, how to ask about touch, and rehearsal of saying or indicating "no". Written notes of participants' touch preferences were made available to facilitators during sessions; practical items such as a buckwheat pillow and blankets were provided to create a sense of weight or grounding. This report uses a case-based, iterative approach rather than a formal quantitative analysis. The investigators reviewed two specific touch events that occurred during their third retreat, evaluated participant-reported outcomes via structured post-session questions and follow-up phone debriefs, and used team deliberations to revise their guidelines. The evolved guidance and its application in a later retreat (the fourth) are presented; the extracted text does not describe predefined analytic statistics or a systematic coding protocol beyond team review practices.

During the third retreat two instances of touch occurred that exceeded the initial guidelines, prompting review and guideline revision. In Case A, a 60-year-old woman with metastatic cancer developed noncardiac chest tightness about 3 hours after psilocybin and, having previously indicated she welcomed touch, placed her hands on her chest. A facilitator placed a buckwheat pillow on her chest and the participant subsequently asked the facilitator to put their hand on top of hers; the facilitator did so and the participant later positioned the facilitator's hand directly over her sternum and kept it there for about 20 minutes. The chest discomfort resolved over the next hour. On the structured day-after question the participant reported no confusing or unwanted touch; in a phone debrief she described the touch as "a lifeline" and, 4.5 months later, characterised it as an "absolute positive experience" that provided connection, grounding and validation. Case B involved a 50-year-old woman with metastatic cancer who began vigorous shaking and later acute distress related to rumination about ancestral harms. Facilitators initially supported her with proximity, encouragement to breathe and extra blankets; when the participant was emerging and remained agitated a certified Reiki practitioner among the facilitators placed hands on the participant's shoulder and hip while coaching breathing. The participant calmed within minutes, moved into a different phase, and later reported no confusion or unwanted touch on the day-after structured question. In phone debriefing she said the touch was "awesome" and later, at 4.5 months, indicated she would request similar touch in the future. Team review concluded that Case A involved participant-initiated touch (consistent with consent) while Case B involved facilitator-initiated touch without in-the-moment explicit consent, raising concern. Nevertheless, the facilitators judged that Case B's distress posed a risk of harm and that the touch likely prevented further distress. The structured and follow-up participant feedback suggested both instances were experienced positively, but the team decided the initial guidelines were insufficient to address such situations. Key revisions included attention to nonverbal signalling (shifting from the term "preferences" to "reactions"), introducing an explicit safe word or gesture (for example "stop", head shake, or thumbs down), and adding body-based practices such as draping. The team also required that facilitators who administer touch beyond usual hand/arm/foot contact should have in-person body-based training with supervised practice and that every instance of "considered" touch be reviewed post-session. The evolved iteration of guidelines was applied in the fourth retreat after additional facilitator training. In that retreat all six participants provided informed discussion-based consent that included the possibility of "considered" touch, and all six answered "no" to the structured day-after question about confusing or unwanted touch. Two participants gave specific comments: one described a moment of ambiguous hand movement onto a blanket as unclear but ultimately helpful, and a male participant appreciated an initiated fist bump from a male facilitator.

Back and colleagues interpret their experience as exposing an ethical tension between a highly circumscribed touch policy and facilitators' clinical judgements in acute situations. They argue the initial, restrictive rules did not adequately allow for participant-initiated contact or for interventions intended to reduce imminent distress, and that team reflection identified the need for a framework that permits limited, carefully controlled deviations when warranted. The evolved guidelines distinguish "usual" touch (hands, arms, feet; routine monitoring) from "considered" touch (intentional contact beyond usual areas to ground or ensure safety) and embed safeguards: pre-session verbal consent conducted in an ordinary state of consciousness, no reconsent once the psilocybin session has begun, the presence of two facilitators for considered touch, draping of the contacted body area, in‑the‑moment seeking of acknowledgement (verbal or nonverbal), trial withdrawal of touch after 2–3 minutes, requirement for facilitators to have body-based training and supervised practice, and routine post-session review and accountability. The authors frame these measures as balancing participant safety, therapeutic support and boundary protection while recognising the power of touch as a nonverbal modality. The paper notes clear limitations acknowledged by the study team. The guidelines were developed in the context of a group retreat intervention for patients with metastatic cancer and have not been rigorously tested; the authors call for larger samples, systematic facilitator reporting, and participant-reported outcomes to better evaluate safety and efficacy. They also recognise that the participants who provided feedback may not represent populations with more extensive trauma histories. The investigators present their evolved guidelines as a work-in-progress rather than universal standards and recommend further empirical study of haptic experience, reciprocity, formality and dominance, as well as routine incorporation of case reports of harms into guideline development.