Psilocybin-assisted therapy for demoralisation in hospice patients: feasibility, safety and preliminary efficacy

Beaussant and colleagues situate this study within evidence that psilocybin-assisted therapy (PAT) can reduce depression, anxiety and demoralisation in people with serious illness, noting that prior trials predominantly enrolled patients with relatively preserved functional status. They argue that terminally ill patients receiving hospice care experience particularly intense psycho-existential distress—commonly labelled demoralisation—which is distinct from major depression and characterised by hopelessness, loss of meaning and a thwarted desire to cope. The introduction highlights that hospice patients have been largely excluded from prior research and that their rapid clinical decline, emotional vulnerability and reliance on interdisciplinary and family-based care raise specific logistical and ethical questions about PAT implementation. This study therefore aims to evaluate the feasibility, safety and preliminary efficacy of a manualised PAT protocol for demoralisation in patients enrolled in home hospice. The investigators designed a prospective, open-label pilot trial to test whether PAT can be integrated into hospice care, assess recruitment and retention, describe adverse events, and measure changes in demoralisation and related psychosocial outcomes, with special attention to real-world operational challenges and patient acceptability.

This was an open-label pilot trial conducted at a large non-profit community hospice in Massachusetts between March 2022 and January 2024. Weekly screenings of the entire home hospice census identified potential participants; the interdisciplinary team (IDT) reviewed candidates before approach, and informed consent was obtained with ongoing re-evaluation. The trial enrolled adults aged 21 or older who were on home hospice, had a Palliative Performance Scale score of at least 40%, moderate-to-severe demoralisation (Demoralisation Scale II, DS-II, score ≥8) and a caregiver present the night of dosing. Key exclusions included current General Inpatient hospice status, personal or family history of psychosis and potential adverse drug–drug interactions. A full list of criteria was available on the trial registry. The intervention combined a manualised psychotherapeutic approach with a single oral dose of 25 mg cGMP psilocybin. Psychotherapy drew on an existential, patient-centred manual adapted from Usona's facilitator manual and incorporated guided imagery and music elements; sessions were delivered by a dyad (lead facilitator—doctoral-level psychotherapist or physician—and a co-facilitator with at least a bachelor's in a mental health field). Participants completed two home-based preparatory sessions, underwent an 8-hour dosing session at the inpatient hospice house with curated music and continuous facilitator support, and received two home-based integration sessions on day 1 and day 7. The hospice IDT collaborated closely with the therapy team to ensure continuity of care. Safety monitoring included hourly vital signs during dosing, assessment of challenging experiences using the Challenging Experience Questionnaire (CEQ) on the dosing day, longitudinal tracking of adverse events (AEs) and serious adverse events (SAEs) up to 24 weeks using CTCAE v5 criteria, suicide risk screening with the Columbia-Suicide Severity Rating Scale (C-SSRS) at each visit and delirium screening with the Confusion Assessment Method. Attribution of AEs to psilocybin or psychotherapy was based on timing relative to drug effects. Efficacy assessments occurred at baseline, week 1, week 3 and monthly up to 24 weeks. The primary clinical outcome was demoralisation measured by the DS-II (16 items, two subscales: meaning and purpose; distress and coping ability). Secondary measures included FACIT-Pal (global quality of life), Pain Interference Scale, HADS-A and HADS-D, LAP-R (death acceptance), Social Isolation Scale, FACIT-Sp (spiritual well-being) and MEQ-30 (mystical experience). Acceptability was measured with the Reaction to Research Participation Questionnaire (RRPQ). Semi-structured qualitative interviews were conducted post-intervention; qualitative findings are reported separately. Statistical analysis was descriptive: only dosed participants were analysed for safety and efficacy, continuous changes were summarised with means and SDs, paired comparisons used Wilcoxon signed-rank tests, and Cohen's d (mean change divided by baseline SD) was reported as a standardised effect-size estimate. No formal repeated-measures modelling was performed due to the small sample.

Screening and enrolment: Over 22 months, 4,607 home hospice patients were screened; 66 were given a study flyer (1.4% of the population), 35 were contacted by a study physician and 15 enrolled (23% of those approached). Ten participants (67% of enrolled) received psilocybin and were included in safety and efficacy analyses; nine (60% of enrolled) completed the week 1 assessment. The mean time from screening to enrolment was 36.6 days (range 2–107), and mean time from enrolment to dosing among treated participants was 13.2 days (range 7–20). Six participants required tapering of serotonergic antidepressants before dosing. Participant characteristics: Treated participants (n=10) had a mean age of 62.7 years (range 47–81, SD 11.3), 50% were female, all identified as White, 50% were married and 70% had at least college education. Seven had cancer (50% gastrointestinal), three had chronic lung disease. Six participants were taking opioid therapy, with a mean daily oral morphine equivalent reported as 485 mg (range 12–1,755 mg, SD 659.9). Half reported current cannabis use for symptom management, 60% prior psychotherapy experience and 50% prior psychedelic use more than 10 years earlier. Feasibility and acceptability: Only a small proportion of the hospice census met eligibility and were approached, reflecting rapid clinical decline and other exclusions such as cognitive impairment, CNS involvement or cardiovascular disease. Of those approached, 15 enrolled and willingness to participate appeared high among eligible patients. The pre-specified acceptability benchmark (80% of completers agreeing or strongly agreeing with selected global RRPQ items) was not met, but the mean total RRPQ score was 90.1/115, indicating generally favourable attitudes. Three participants rated some global items neutrally despite endorsing personal value and perceived respect. Safety: No SAEs were attributed to psilocybin. Reported treatment-related AEs (hypertension, nausea, headache) were generally mild to moderate and expected. Two participants (20%) required overnight observation post-dosing—one for prolonged effects and grade 3 nausea, one for fatigue. Two transient asymptomatic grade 3 hypertension episodes occurred during dosing; one diabetic participant had transient asymptomatic hypoglycaemia six hours after administration. No suicidality was reported on the C-SSRS, and there was no clinical worsening of oxygen requirement during dosing among participants on supplemental nasal oxygen. Primary efficacy—demoralisation: Despite ongoing clinical decline, the DS-II total score decreased significantly from baseline to week 3 by a mean of 8.8 points (95% CI 2.25 to 14.75, p=0.0196). Changes at week 1 (mean −5.3, 95% CI −7.02 to 16.98, p=0.0784) and week 7 (mean −2.7, 95% CI −2.07 to 27.13, p=0.624) were not statistically significant. At week 3 (n=9), five participants (55%) met the study definition of treatment response on demoralisation (≥30% reduction), and three of these were in remission (DS-II <8). Subscale findings showed a significant reduction in the meaning and purpose subscore only at week 3 (p=0.0487), while the distress and coping ability subscore was significantly improved at both week 1 (p=0.0429) and week 3 (p=0.0183). Secondary and exploratory outcomes: Improvements were observed across several secondary measures (FACIT-Pal, Pain Interference Scale, HADS-A/D, LAP-R, Social Isolation Scale, FACIT-Sp), but most did not reach statistical significance. The FACIT-Sp total score increased by 8 points at week 3 (p=0.0566), narrowly missing conventional significance; its peace subscale showed a statistically significant increase (p=0.00566). Acute subjective effects: The CEQ mean score was 36.6 (range 12–66, SD 20.8), equating to 28% of the maximum; grief-related items constituted the largest proportion (46%) of CEQ scoring, followed by physical distress (33%) and fear (25%). Seven participants rated at least one grief item as strong or extreme. The MEQ mean score was 84.1 (range 46–137, SD 34.47), corresponding to 56% of the maximum; ineffability contributed about 63% of MEQ scores, and seven participants endorsed strong or extreme items across mystical, transcendence or positive mood domains. Illustrative qualitative narratives (reported separately) reflected both mystical and challenging aspects including grief, transcendence and fear.

Beaussant and colleagues interpret these data as the first prospective evidence that PAT can be feasibly and safely delivered to a carefully selected subset of terminally ill hospice patients and that it may reduce demoralisation in this context. They emphasise that the statistically significant and clinically meaningful reduction in DS-II at week 3 is notable given participants' ongoing disease progression and symptom burden; the finding suggests PAT may facilitate emotional and existential processing near the end of life. The authors also stress that the intervention was feasible only for a small fraction of the screened hospice population (1.4%), largely because of rapid clinical decline, cognitive impairment and logistical barriers, underscoring constrained generalisability and operational challenges. The discussion highlights grief as a prominent acute challenge during dosing in this population, with grief-related items dominating CEQ responses, while mystical-type experiences were present but, on average, lower than in prior cancer studies. The significant increase in the FACIT-Sp peace subscale alongside reductions in demoralisation is presented as consistent with meaningful existential effects, but the authors underline that the intensity of these experiences may be inappropriate for some patients with limited energy or time. They further argue that the benefits observed likely depend on integration within high-quality interdisciplinary hospice care rather than on psilocybin as a standalone treatment. Practical lessons and suggestions for future research include the adequacy of two preparatory sessions in this setting, the potential value of additional optional integration sessions (for example at weeks 2 and 3), and consideration of flexible dosing strategies or shorter-acting alternatives such as ketamine to broaden accessibility. The authors candidly note limitations: small sample size, lack of a control group, homogeneous demographics, missing data at later time points, descriptive statistical approach without adjustment for multiple comparisons and exclusion of patients with primary cardiac or hepatic disease. They present this work as hypothesis-generating, intended to inform the design of larger controlled trials and to refine feasibility and acceptability metrics for hospice settings.

The investigators conclude that the PATH pilot trial provides initial prospective evidence that psilocybin-assisted therapy can be safely and meaningfully delivered to a subset of terminally ill patients receiving hospice care. Despite significant clinical and logistical hurdles, the intervention was generally well tolerated and associated with a notable reduction in demoralisation at three weeks. Although the pre-specified acceptability benchmark was not fully met, overall RRPQ ratings were high, indicating that participants found the experience respectful and valuable. The authors call for future studies to refine feasibility criteria and to develop adaptive, patient-centred models that optimise safe and ethical integration of psychedelic therapies into hospice and palliative care.