Psilocybin-Assisted Group Psychotherapy + Mindfulness Based Stress Reduction (MBSR) for Frontline Healthcare Provider COVID-19 Related Depression and Burnout: A Randomized Clinical Trial

Depression and burnout among physicians and nurses are described as worsening crises within the US healthcare system that were amplified by the COVID-19 pandemic through elevated clinical demand, resource constraints and increased personal risk. Burnout is framed as a syndrome of emotional exhaustion, depersonalisation and reduced personal accomplishment that can impair clinician–patient relationships. Mindfulness training, particularly the established Mindfulness-Based Stress Reduction (MBSR) programme, is presented as an evidence-based behavioural intervention that can reduce symptoms of depression, anxiety and burnout in patients and healthcare providers. Separately, psychedelics such as psilocybin have shown efficacy for depressive symptoms, and there is growing interest in potential synergy between mindfulness training and psychedelic-assisted psychotherapy (PAP), based on overlapping neural mechanisms and the idea that psychedelic experiences may deepen mindfulness skills and produce more durable change when combined with formal training. Lewis and colleagues set out to evaluate whether adding group-format psilocybin-assisted psychotherapy to an 8-week MBSR curriculum is safe, feasible, and more effective than MBSR alone for frontline physicians and nurses with COVID-19–related depressive symptoms and burnout. The trial aimed to examine feasibility and safety outcomes, with the primary clinical endpoint defined as change in depressive symptoms at two weeks post-intervention using the QIDS-SR-16, and multiple secondary outcomes including burnout subscales, demoralization and measures of connectedness. The study also explored expectancy and experiential measures as potential moderators or correlates of outcomes.

This was a randomised, controlled trial that enrolled frontline physicians and nurses who reported depressive symptoms and burnout related to their COVID-19 clinical work. After screening procedures (including metabolic panel, urine drug screen and pregnancy test where applicable) and baseline assessments, participants were randomised 1:1 in blocks of 3–5 per cohort to either MBSR alone or MBSR plus group psilocybin-assisted psychotherapy (MBSR+PAP). Allocation was generated by an investigator not involved in assessment or analysis and assigned by sealed envelope. Because no placebo drug was used, participants were not blinded to psychedelic treatment; the statistician remained masked to allocations until study completion. All participants attended a standard 8-week MBSR course consisting of eight weekly two-hour group sessions teaching mindfulness meditation (for example, mindful breathing and body scan) and psychoeducation. For those randomised to MBSR+PAP, the psilocybin intervention began after four weeks of MBSR and comprised three group preparatory sessions over one week, a single group dosing session following established group PAP protocols, and three group integration sessions over the subsequent two weeks. Each participant in the MBSR+PAP arm was paired with an individual therapist and received a 30-minute one-to-one break-out during preparatory and integration sessions; the group therapy followed a supportive–expressive model and therapists were supportive and non-directive during dosing. Vital signs were monitored during dosing and adverse event (AE) and suicidality assessments were completed before participants left the dosing site. Participants in the MBSR-only arm attended an in-person all-day silent meditation retreat timed to coincide with the psilocybin dosing day for the other arm. Attendance targets for feasibility were specified as at least 66% attendance at MBSR sessions for both arms and 75% attendance at preparatory and integration sessions for the MBSR+PAP arm. Primary clinical outcomes were assessed at baseline, 2-weeks and 6-months post-intervention. The primary endpoint was change in depressive symptoms measured by the Quick Inventory of Depressive Symptomatology–Self Report (QIDS-SR-16) at 2 weeks; this is a 16-item self-report instrument with total scores ranging 0–27, where higher scores indicate more severe depression and a 3.5-point reduction was treated as clinically meaningful. Key secondary outcomes included the Maslach Burnout Inventory Human Services Survey for Medical Professionals (MBI-HSS-MP) with subscales for emotional exhaustion, depersonalisation and personal accomplishment, the Demoralization Scale (DS-II), and the Watts Connectedness Scale (WCS) measuring connectedness to self, others and world. Expectancy and experiential measures (Credibility/Expectancy Questionnaire, Mystical Experience Questionnaire MEQ-30, Challenging Experience Questionnaire CEQ, and NADA-state) were collected pre- and post-randomization or after the dosing/retreat day as appropriate. The trial planned sample size on prior meta-analytic effect sizes for psilocybin on depression; anticipating 18% dropout, an assumed effect size around 1.0 and α=0.05, the target was approximately 24 participants to achieve 80% power. Analyses used intent-to-treat (ITT) including all randomised participants (N=25) and a per-protocol sample (N=20) defined by attendance thresholds. Linear mixed models (LMMs) with maximum likelihood estimation and random intercepts were used to evaluate treatment-by-time interactions; time was treated categorically. Correlation analyses assessed the role of post-randomisation expectancy on QIDS change, and linear regressions summarised relationships between experiential measures and outcomes. Analyses were performed in R 4.4.0 with significance set at α=0.05.

Of 420 people screened, 25 participants were enrolled and randomised. Baseline characteristics in the ITT sample included mean QIDS-SR-16 score 12.3 (SD 3.9), indicating moderate depression, 72% female, 96% white, and 10 MDs and 15 RNs; 88% reported lifetime antidepressant use. Mean ages were 40.4 (SD 8.4) in the MBSR-only arm and 47.4 (SD 10.9) in the MBSR+PAP arm. Three participants withdrew for scheduling reasons (one from MBSR+PAP, two from MBSR-only); no withdrawals were due to adverse events. Per-protocol analyses excluded two participants who did not attend at least two-thirds of scheduled MBSR sessions, yielding N=20. Feasibility targets were met or exceeded: preparatory sessions attendance in MBSR+PAP was 100%, the dosing session attendance was 100%, integration session attendance was 97.2%, and overall scheduled MBSR session attendance across both arms was 80.9%. Adverse events were collected at multiple time points through 6 months using CTCAE v.5 and suicidality was monitored with the C-SSRS. The extracted text reports 12 study-related adverse events that were Grade 1–2 and states there were no serious adverse events or emergent suicidality through follow-up. For the primary clinical outcome, MBSR+PAP produced significantly greater reductions in depressive symptoms at the 2-week primary endpoint compared with MBSR-only: between-groups effect = 4.6 points on the QIDS-SR-16 (95% CI 1.51–7.70), p=0.004, Cohen's d=1.04. On burnout subscales, the MBSR+PAP arm showed greater reduction in depersonalisation from baseline to 2 weeks (between-groups effect = 5.47, 95% CI 0.3–10.6, p=0.038), and greater reduction in emotional exhaustion from baseline to 6 months (between-groups effect = 10.9, 95% CI 2.1–19.8, p=0.016). There were no significant between-group differences on the personal accomplishment subscale, attributed to ceiling effects at baseline. Regarding demoralization and connectedness, MBSR+PAP outperformed MBSR-only in reducing demoralization at the 2-week endpoint. The WCS General Connectedness measure favoured MBSR+PAP from baseline to 2 weeks (between-groups effect = -17.5, 95% CI -29.6 to -5.5, p=0.005), with significant differences on the Connectedness to Self and Connectedness to Others subscales as well. Significant treatment-by-time interactions across baseline, 2-week and 6-month time points were reported for QIDS-SR-16, MBI emotional exhaustion, WCS General Connectedness and WCS subscales. Expectancy ratings differed markedly after randomisation: mean post-randomisation expectancy was higher in the MBSR+PAP arm (65.4, SD 14.7) than MBSR-only (37.6, SD 17.9), p=0.0003. Expectancy correlated strongly with QIDS change in the MBSR-only arm (r=-0.70, p=0.022) but not in the MBSR+PAP arm (r=0.04, p=0.90). Across the full sample, magnitude of mystical-type experience (MEQ-30) and NADA-state scores correlated with greater reductions in QIDS-SR-16 from baseline to 2 weeks (r=-0.62, p=0.0019 for MEQ-30; r=-0.65, p=0.0018 for NADA-state). MEQ-30 scores also correlated with changes in burnout emotional exhaustion (r=-0.47, p=0.0286), depersonalisation (r=-0.44, p=0.0421), and WCS (r=0.616, p=0.0023). There were no significant correlations between CEQ scores and change in primary or secondary outcomes at 2 weeks.

Lewis and colleagues interpret their findings as evidence that a single 25 mg group psilocybin dosing session delivered within an 8-week MBSR course is feasible, safe in this sample, and associated with clinically and statistically significant short-term reductions in depressive symptoms and burnout compared with MBSR alone. The trial reported no serious treatment-emergent adverse events and no emergent suicidality or self-injurious behaviour through follow-up. The authors emphasise a large antidepressant effect at the 2-week endpoint and parallel increases in connectedness to self and others, which they link to theoretical and empirical literature implicating social connection in both the development and amelioration of depression and burnout. The group delivery model is highlighted as intentionally addressing social factors relevant to clinicians and as a potentially scalable format compared with individual PAP approaches that typically use high therapist-to-participant ratios. The authors note that experiential measures, notably the MEQ-30, were strongly correlated with clinical improvement across the whole sample, and they suggest this supports the idea that self-transcendent states—whether occasioned by psilocybin or by intensive mindfulness practice—may contribute to therapeutic change. Key limitations acknowledged include the small sample size and the homogeneity of participants (predominantly white women), which constrain generalisability. The open-label design and lack of an active placebo control are cited as limitations that may permit expectancy or other non-specific effects; indeed, post-randomisation expectancy differed between arms and was associated with outcomes in the MBSR-only group. The interventions were not equivalent in experiential content (psilocybin dosing day versus silent retreat), and prior psychedelic experience was not an exclusion criterion (six participants per arm had prior use), leaving uncertainty about differential effects by prior exposure. The authors recommend future double-blind RCTs with active placebo controls or full factorial designs to disentangle independent and interactive effects of psilocybin and mindfulness training, and they call for larger, more diverse, multi-site studies to confirm safety and efficacy and to better characterise mechanisms and moderators.