A Systematic Review of Reporting Practices in Psychedelic Clinical Trials: Psychological Support, Therapy, and Psychosocial Interventions

Psychedelic-assisted therapies (PAT) combine a pharmacological agent (classic psychedelics or MDMA) with accompanying psychosocial interventions (PI). Prior empirical and historical sources suggest that relational and psychotherapeutic elements—therapeutic alliance, preparatory work, integration, and other forms of non-drug support—can shape subjective experiences and clinical outcomes. Evidence cited by the authors includes trial-derived analyses linking therapeutic alliance and emotional breakthrough to reductions in depressive symptoms, experimental comparisons of high- versus low-support adjuncts to psilocybin, observational studies of retreat participants that connect facilitator rapport to sustained well-being, and historical trials from the 1960s that suggest trials with more comprehensive psychosocial input yielded superior outcomes. Qualitative reports from trial participants further emphasise safety and rapport as central to perceived benefit, while safety literature raises concerns that inadequate PI may contribute to adverse events or harmful dynamics. B. and colleagues set out to evaluate how well PI are reported in contemporary PAT clinical trial publications. Their goal was practical: to assess whether trial reports provide enough detail to allow replicability in clinical practice and to permit assessment of generalisability. To do this they conducted a systematic review of quantitative clinical trials published in English since 2000 and assessed trial reports against a modified set of reporting items derived from established guidance (CONSORT/TIDieR) and prior work on multicomponent interventions. The review focused on reporting practices for non-drug psychosocial components rather than on trial efficacy per se.

The review targeted primary publications of quantitative clinical trials that treated patients with a psychiatric indication using either classic psychedelics (for example psilocybin, LSD, DMT, ayahuasca) or MDMA, published in English from 2000 to May 2023. Included designs comprised open-label studies and randomised controlled trials; excluded were reviews, surveys, secondary analyses, case reports, and studies of healthy volunteers. The authors specified these criteria to capture publications reporting primary clinical efficacy data for PAT interventions. On 26 May 2023 the researchers ran systematic searches in PubMed/Medline and PsycINFO using two concurrent search strings (one listing psychedelic compounds and the other focusing on clinical-trial study types), adapting terms from previous related work. They corroborated the search by checking reference lists of reviews with similar inclusion criteria. The extracted text refers to supplemental materials for the precise search terms and to a figure showing the screening flow, but those items are not included in the provided extraction. To evaluate reporting completeness the study used a modified version of a 19-item assessment developed by Candy and colleagues, which in turn was based on CONSORT extension and the TIDieR checklist. The authors assessed 16 of the original items (excluding some that they judged inapplicable) and added two items specific to PAT reporting: whether a therapy manual was available and a qualitative assessment of the descriptor used for non-drug sessions. Each assessed item (except the qualitative descriptor) was coded dichotomously as adequately reported (1) or not (0). The search for reporting included both primary publications and any associated supplementary materials. Data extraction was performed by a single author and independently checked by two other authors, with discrepancies resolved through discussion. The investigators did not perform formal risk-of-bias assessment because it was not central to their objective of evaluating reporting practices. Where feasible, the authors compared their findings with reporting frequencies reported by Candy et al. for non-psychedelic multicomponent interventions, while noting several caveats about differences in scope, search strategy, and publication dates.

The database searches yielded 1,142 records (PubMed n = 1,049; PsycINFO n = 93). After duplicates were removed, 1,078 titles and abstracts were screened, leading to full-text assessment of 45 articles. Twelve full texts were excluded (11 secondary analyses and 1 preliminary report), leaving 33 primary trial reports included in the review. Risk of bias for the included trials was not assessed because the study's focus was on reporting quality rather than internal validity. Across the 33 included trials the authors found widespread underreporting of psychosocial intervention details. Reported frequencies (presented by the authors in the Discussion but representing extracted results) included: 42% of trials did not report provider credentials; 52% did not report whether a therapy manual was used; 67% did not reference a therapy manual that was available to readers; 33% did not report the number of non-drug sessions; 45% did not report duration of sessions; 82% did not report assessing treatment fidelity; and 100% did not report fidelity outcomes. When directly compared to available data from multicomponent non-psychedelic trials, PAT reports lagged the comparator by at least 10 percentage points on six of 16 items: treatment setting, provider credentials, number of sessions, intervention supporting materials, tailoring to participants, and fidelity outcome. By contrast, psychedelic reports outperformed the comparator by at least 10 percentage points on a single item, “descriptive aids.” The authors highlight specific recurrent deficiencies that lowered aggregate reporting scores: routine omission of details about non-drug sessions (frequency, timeline, duration), failure to document preparatory and integration session schedules, unreported ad hoc additional sessions or supports and unclear criteria for providing them, inconsistencies between information in primary papers and supplementary materials (for example differing session counts or nomenclature such as "debrief" versus "integration"), references to established manualised therapies without clear description of adaptations for PAT, and reliance on the presumed standardisation of borrowed psychotherapeutic approaches rather than standardising the actual adapted PI used in the trial. The review also noted that many authors used supplementary materials or referenced publicly available manuals to provide some PI detail; on that item PAT reports exceeded non-psychedelic trials, but the extraction could not capture the contents of referenced secondary publications or later manuals not contemporaneous with the primary report.

B. and colleagues interpret their findings as demonstrating substantial and important gaps in the reporting of psychosocial interventions in contemporary PAT trial reports. They argue that omissions—about provider credentials, session number and duration, manuals, and fidelity assessment—undermine reproducibility, the ability to assess what component(s) produced observed effects, and the generalisability of outcomes to clinical settings. The authors situate these concerns against prior empirical and qualitative work that suggests PI elements such as therapeutic alliance, pre-session rapport, and the amount of non-drug support can materially influence participant experience and downstream clinical outcomes. The discussion considers plausible contributors to the reporting deficiencies. One explanation offered is publication convention: many PAT trial reports appear in journals prioritising pharmacotherapy, which may emphasise drug-related methods and leave limited space for detailed psychotherapy description. Authors sometimes mitigate this by using supplementary materials, accounting for why PAT reports scored higher on that particular item. The investigators also suggest systemic incentives that downplay psychosocial complexity—commercialisation pressures to minimise projected treatment cost, and trial designs that intentionally hold PI constant while varying drug conditions—may reduce attention to documenting PI. They caution that omission of PI details could have practical consequences if future trials or clinical implementations reduce or omit psychosocial supports without evidence, potentially affecting safety and efficacy. To address the identified shortfalls the authors propose a concise set of reporting recommendations (derived from TIDieR observations) intended to augment existing guidelines and target common PAT deficits; the extraction notes these are listed in a table but does not provide the table content. The study acknowledges limitations that temper interpretation: the search was limited to English-language databases and may miss non-English trials; search terms and chosen databases might have overlooked relevant studies; data extraction was carried out by a single author before independent review, introducing potential subjectivity; and the quantitative coding could not adequately capture the use of separate secondary publications or later manuals that some teams referenced to describe PI. The authors conclude by calling for improved standardisation and fuller reporting of PI to support replication, safety assessment, and implementation of PAT.

The authors conclude that reporting of psychosocial interventions in psychedelic clinical trials is often incomplete. Many primary reports lack essential details—number and duration of non-drug sessions, provider qualifications, availability of manuals, and fidelity assessment—which complicates reproducibility, assessment of efficacy, and translation of PAT into clinical practice. Compared with reporting in non-psychedelic multicomponent trials, PAT reports underperformed on several key items. To address this gap the authors recommend that future PAT publications include a minimum set of reporting elements (they refer to nine suggested elements) and argue that comprehensive reporting is necessary to standardise research, safeguard participant safety, and improve treatment outcomes.