Introduction

Clinical research with psychedelics has provided promising positive results for various mental disorders (Andersen et al., 2020; Bahji et al., 2020; Dos Santos et al., 2018; Vargas et al., 2020). However, sample sizes were generally small and rather selective. With the first phase III randomized clinical trials (RCTs) nearing completion (Mitchell et al., 2021) and many trials currently investigating novel applications (Siegel et al., 2021), some compounds seem close to registration.

The use of psychedelics, particularly in uncontrolled circumstances, is associated with acute adverse events (AEs) such as anxiety, panic, dysphoria, paranoia, and/or dangerous behaviors (Barrett et al., 2016; Johnson et al., 2008); such events may contribute to enduring psychological problems (Carbonaro et al., 2016). Although adverse outcomes were mostly described after non-medical psychedelic use, and safety and tolerability have been demonstrated in small clinical trials (Andersen et al., 2020; Dos Santos et al., 2018; Rucker et al., 2018), administration in larger, more heterogeneous patient populations with higher levels of comorbidity may lead to unexpected negative effects. Moreover, adverse drug reactions may lead to non-adherence and discontinuation of treatment (Carvalho et al., 2016), and unresolved and non-integrated difficult experiences may lead to persisting negative outcomes (Grof, 2001; Johnson et al., 2008). It is therefore important to identify the full range of adverse reactions to psychedelic drugs, particularly in vulnerable patients with treatment-resistant mental disorders. The literature on this topic has not been systematically described since 1984 (Strassman, 1984). A 1960 review, based on questionnaires distributed among clinicians representing some 5000 patients (mostly treated with lysergic acid diethylamide (LSD)), concluded that AEs were rare, with some exceptions, mostly in patients with schizophrenia, and that these drugs were generally safe when administered with care (Cohen, 1960). A 1984 review of adverse reactions to psychedelics in different settings found that AEs existed on a continuum; from acute, time-limited panic reactions during administration, through transient psychoses lasting several days, to recurrent flashbacks and chronic undifferentiated psychotic and treatment-resistant cases (Strassman, 1984). Flashbacks were later included in the diagnostic category “hallucinogen persisting perception disorder” (HPPD, which appears in the Diagnostic and Statistical Manual of Mental Disorders (DSM-V) but not in the International Statistical Classification of Diseases and Related Health Problems (ICD-10); HPPD is reported occasionally, mostly in the context of frequent non-medical use of serotonergic psychedelics or 3,4-methyenedioxymethamphetamine (MDMA), and may be related with pre-existing psychiatric conditions (Halpern and Pope, 2002; Halpern et al., 2018; Litjens et al., 2014); flashbacks seem to be mostly perceived as mild and neutral to pleasant (Müller et al., 2022). A recent non-systematic, narrative review explored the evidence base for the most frequently mentioned AEs in public discourse, to elucidate which of these harms are based largely on anecdotes versus those that stand up to current scientific scrutiny (Schlag et al., 2022).

Assessing AEs is challenging for multiple reasons; AEs are not always pre-specified, the range of potential reactions can be broad, reporting can be erratic, and terminology is often inconsistent (Golder et al., 2016, 2019) and includes side effects, toxic effects, adverse effects, treatment-emergent AEs (TEAEs), adverse drug reactions, complications and harms, all of which are used interchangeably (Peryer et al., 2021). This is further complicated by the highly variable and context-dependent subjective effects elicited by psychedelics (Breeksema et al., 2020; Carbonaro et al., 2016).

As more patients with mental disorders will probably be treated with psychedelics in the near future, a complete overview of both therapeutic benefits and AEs is needed for balanced benefit-risk assessments and decisions, and for understanding which patients are most likely (not) to profit (Loke et al., 2007). The current paper aims to systematically review any AEs occurring during or after psychedelic treatments with classic/serotonergic hallucinogens (psilocybin, LSD, ayahuasca) and entactogens (MDMA) in patients. Despite relevant pharmacological distinctions, both MDMA and serotonergic hallucinogens are often classified as “psychedelics” and administered under similar therapeutic conditions (Garcia-Romeu et al., 2016; Reiff et al., 2020). AEs of the atypical psychedelics ketamine (Short et al., 2018; Van Amsterdam and Van Den Brink, 2021) and ibogaine (Ona et al., 2022) have recently been reviewed elsewhere.

Methods

Given the relative novelty of this field, this review takes an exploratory—rather than a confirmatory—approach. The review was not preregistered. This systematic review and data synthesis follows a sequential explanatory design, using the Cochrane Handbook for Systematic Reviews (Peryer et al., 2021), Adverse Effects Subgroup framework (Loke et al., 2007), to extract, organize, and synthetize findings from both quantitative and qualitative studies, following Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines (Moher et al., 2009). We will use the term AE for any unfavorable or harmful event first occurring during or after the administration of a classic psychedelic or MDMA and irrespective of its relationship with the psychedelic treatment (Loke et al., 2007; Peryer et al., 2021). In this review, acute AEs refer to AEs occurring during or on the day of a psychedelic/MDMA session, whereas late AEs refer to AEs that emerge after the day of the psychedelic/MDMA session. “TEAEs” are sometimes used, defined as an AE that is either not present prior to the treatment or an already present event that has worsened following treatment. As there is no clear distinction between these terms, in this review, we will only distinguish between acute and late AEs.

Selection criteria

All quantitative and qualitative clinical studies describing the treatment of patients with a mental disorder with a classic serotonergic psychedelic (e.g., psilocybin, LSD, mescaline, ayahuasca) or an entactogenic drug (e.g., MDMA, MDA), including open-label studies, case reports or case series, and randomized controlled trials (RCTs) published since 2000, were included. We excluded systematic reviews, surveys, secondary analyses, and studies with healthy volunteers. We aimed to assess the possible adverse reactions occurring in patients with a diagnosed (mental) disorder who were treated with any of these compounds. As such, we formulated no specific exclusion criteria a priori to allow the broadest inclusion possible and not miss any potential unfavorable outcome.

Search methods

We systematically searched the PubMed/Medline, EMBASE, and PSYCinfo databases on 28 July 2021, combining search strings containing both index¹ and free-text terms: one group for psychedelic compounds (e.g., psilocybin OR “lysergic acid diethylamide” OR MDMA) AND another for study type (e.g., “clinical trial” OR qualitative), excluding animal studies. See Supplemental Material 1 for a detailed list of search terms. In addition, we searched ClinicalTrials.gov to cross-check corresponding publications. Systematic searches were complemented by handsearching and checking reference lists.

Quality assessment

A thorough formal quality assessment was not considered pertinent since high-quality, low bias studies may still report poorly on AEs (Loke et al., 2007). We specifically assessed quality criteria relevant in the context of the aim of this review: methods used to monitor or report AEs, study inclusion and exclusion criteria, and the percentage of participants with prior experience with the drug. The Critical Appraisal Skills Programme (CASP) checklist was used to assess the methodological rigor of qualitative studies (Critical Appraisal Skills Programme, 2020).

Data extraction and synthesis

We extracted all descriptive and quantitative information on acute and late AEs, the timing of assessment, and serious AEs (SAEs) from the main text and the supplementary files of the publications. The AE prevalence was recalculated as a percentage of total participants. Qualitative articles were scrutinized for themes and descriptions related to AEs. These were then used to complement, illustrate, and contextualize findings from quantitative studies.

Results

We found 5640 articles (PubMed, n = 1832; EMBASE, n = 2735; PsycINFO, n = 1073). After removal of duplicates, the remaining 3709 titles and abstracts were screened independently by the first two authors. Any discrepancies on inclusions were discussed between multiple authors until consensus was reached. After screening, the full texts of 61 articles were assessed for eligibility; 44 articles were included. We found 596 registered trials; after selection, 24 completed trials were included (see flow diagram, Figure 1).

Figure 1. Flowchart.

Included studies were published between 2006 and 2021 and reported on a total of 598 unique participants, of whom 521 received an active dose. Sample sizes varied from 1 to 105 participants (mean n = 23), with mean ages ranging from 25 to 59 years. Multiple publications referring to a single study were merged. All qualitative studies described subsamples from quantitative studies included in this review. There was substantial variation in study design, substances, dosages, and disorders. Four quantitative and two qualitative studies did not report on AEs.

Below, we describe the main results per compound, differentiated per disorder. Table 1 provides an overview of included studies, participants, drug used, prior psychedelic use, AE measurements, and a summary of reported acute and late AEs, including SAEs. For ease of reading, here we only present the 3–5 most reported AEs – a detailed overview of all AEs reported is provided in Supplemental Table S1. Table 2 summarizes all included qualitative studies.

Table 1. All 34 included quantitative studies.

Table 2. Overview of all 10 included qualitative studies.

MDMA

In all, 16 studies were selected, including one qualitative study (total N = 266 patients treated). In total, 10 publications described MDMA-assisted treatment of PTSD (total N = 214 patients): seven RCTs (Mitchell et al., 2021; Mithoefer et al., 2011; 2013b, 2018, 2019; Oehen et al., 2013; Ot’alora et al., 2018), one open-label trial (Jardim et al., 2020), one case series (Monson et al., 2020; Wagner et al., 2019), and one qualitative study (Barone et al., 2019). Other RCTs reported on the treatment of end-of-life anxiety (EOLA) (Wolfson et al., 2020), social anxiety in adults with autism (Danforth et al., 2018), and tinnitus (Searchfield et al., 2020); one open-label study focused on alcohol use disorder (AUD) (Sessa et al., 2019; 2021b). Active doses ranged from 50 to 125 mg (sometimes followed by an optional supplemental half-dose).

Quality appraisal

Of all MDMA studies, 11 reported AEs only when spontaneously reported by participants; one study did not report on AEs (Searchfield et al., 2020). Two RCTs systematically assessed AEs (Mitchell et al., 2021; Mithoefer et al., 2019), another study employed the UKU scale of secondary effects; this was the only study reporting AE severity (Oehen et al., 2013). Six studies did not report whether participants had ever used MDMA or “ecstasy” prior to study participation (Bouso et al., 2008; Jardim et al., 2020; Monson et al., 2020; Ot’alora et al., 2018; Searchfield et al., 2020; Sessa et al., 2021b); prior exposure to MDMA in the other nine quantitative studies varied from none (Danforth et al., 2018) to 56% (Wolfson et al., 2020). The qualitative study was of medium/high quality (Barone et al., 2019) but did not report on AEs.

Physiological effects

Nine studies measured blood pressure, body temperature, and heart rate (HR); all studies reported (mild) transient, statistically significant elevations during the MDMA session. None required medical intervention. One MDMA-related SAE was reported, in which a participant experienced an increase in premature ventricular contractions (also present at baseline), which required one night of hospitalized monitoring, and resolved spontaneously afterwards (Mithoefer et al., 2018).

Adverse events

PTSD

The most common acute physical AEs were jaw clenching and/or tight muscles, headaches, nausea, fatigue, and lack of appetite. Other acute AEs included feeling cold, thirst, dizziness, perspiration, restlessness, nausea, and somatic pains. Anxiety was by far the most common psychological AE. Most acute AEs occurred more often in the MDMA than in the placebo groups, with some exceptions (e.g., one RCT found a slightly higher prevalence of anxiety and fatigue in the placebo group) (Mithoefer et al., 2011). Acute AEs were mild to moderate severe in the single study that reported severity (Oehen et al., 2013). There was no dose-AE relation in the two PTSD studies that assessed different active MDMA doses (Mithoefer et al., 2018; Ot’alora et al., 2018). SAEs were reported in two studies: in one study two patients reported suicidal behavior and suicidal ideation leading to self-hospitalization, these only occurred in the control group (Mitchell et al., 2021); in another study, three out of four SAEs were considered unrelated to the treatment (Mithoefer et al., 2018); one patient developed an exacerbation of pre-existing premature ventricular contractions, which resolved without evidence for cardiac disease, and was considered potentially related (Mithoefer et al., 2018).

In a case study, Wagner and colleagues (2019) illustrate how a (pseudonymized) participant experienced these acute adverse reactions, and how they related to the therapeutic process:

Stuart experienced strong emotional reactions, such as crying and grief in the sessions, and did not try to stop or escape those experiences. He also had strong visceral reactions in the MDMA session, including muscle tightening and sweating as he reviewed traumatic memories while ‘inside’. Following this session, Stuart reflected on this experience as follows: “There’s no easy fix—I need to work through the darkness.”

Common late AEs were fatigue, lack of appetite, low mood, insomnia, need for more sleep, increased irritability, headache, difficulty concentrating, and anxiety. In one study, rates of fatigue, insomnia, and headaches were somewhat higher in the control group (Mithoefer et al., 2019). A minority of patients in this group reported some complaints up to 2 months after the final session, including anxiety (17 of 72 patients), depressed or low mood (6 of 72 patients), and panic attacks (4 of 72 patients), with higher rates in the MDMA than the control group.

End-of-life anxiety

A crossover RCT for EOLA reported jaw clenching, thirst, dry mouth, perspiration, and headache as acute AEs. Fatigue, need for more sleep, insomnia, anxiety, jaw clenching, and low mood were the most common late AEs (Wolfson et al., 2020). Except low mood, all AEs occurred more in the MDMA group. Anxiety, depressed mood, and dissociation were reported only in the MDMA group; insomnia was reported in both groups.

Social anxiety

The RCT on social anxiety in adults with autism reported higher rates of acute AEs in the MDMA than in the placebo group, including anxiety, difficulty concentrating, fatigue, headache, and sensitivity to cold (Danforth et al., 2018). Most prevalent late AEs included fatigue, headache, difficulty concentrating, low mood, and the need for more sleep; all late AEs except the latter two were more common in the MDMA group. Two of the eight patients in the MDMA group, and one of four patients in the placebo group reported depressed mood, suicidal ideation, and panic attack in the period following the first drug session.

Tinnitus and AUD

The studies on AUD and tinnitus treatment did not report on AEs.

Psilocybin

We included 20 articles, describing psilocybin treatment in 257 patients: six (crossover) RCTs, five articles describing four open-label studies, one single-group, pseudo-randomized dose-escalation study, and eight qualitative studies. Three RCTs on EOLA (Griffiths et al., 2016; Grob et al., 2011; Ross et al., 2016), two RCTs on MDD (Carhart-Harris et al., 2021; Davis et al., 2020), one RCT on migraine headaches (Schindler et al., 2020), and one dose-escalation study on obsessive-compulsive disorder (Moreno et al., 2006). Open-label studies investigated the treatment of treatment-resistant depression (TRD; Carhart-Harris et al., 2016, 2018), AUD (Bogenschutz et al., 2015), tobacco smoking (Johnson et al., 2014), and demoralization in older AIDS-survivors (Anderson et al., 2020). Active dosages varied from 10 to 30 mg or 0.1 to 0.4 mg/kg of psilocybin per session. Only two out of six studies that administered different doses of psilocybin reported on adverse effect per dosage (Davis et al., 2020; Griffiths et al., 2016); one study described the timing of AEs per session (Carhart-Harris et al., 2016). Qualitative studies described treatment experiences among patients with EOLA (Belser et al., 2017; Malone et al., 2018; Swift et al., 2017), TRD (Kaelen et al., 2018; Watts et al., 2017), and substance use disorder (SUD; Bogenschutz et al., 2018; Nielson et al., 2018; Noorani et al., 2018).

Quality appraisal

Two studies did not report on AEs (Grob et al., 2011; Moreno et al., 2006). Most other studies relied on spontaneous reporting by patient and therapist observations; some (also) used post-treatment interviews, post-day headache ratings, and structured questionnaires, for example, Challenging Experience Questionnaire (CEQ) or States of Consciousness Questionnaire. Three studies reported on AE severity (Anderson et al., 2020; Davis et al., 2020; Johnson et al., 2014). In the studies that assessed dose–effect relationships, one study found a higher prevalence of some psychological AEs in the moderate-dose (20 mg) than in the high-dose (30 mg) session (Davis et al., 2020) and another study reported a higher prevalence of post-treatment headaches in the higher dose group (Griffiths et al., 2016). One study reported a much higher number of distinct AEs than all other studies (Davis et al., 2020).

With one exception (Bogenschutz et al., 2015), all quantitative studies reported on prior psychedelic use, varying from 20% (Schindler et al., 2020) to 100% (Moreno et al., 2006), averaging 50% of patients with psilocybin or other psychedelic use before the study. One study reported a group average of 0.8 lifetime experiences with psychedelics (Davis et al., 2020); in another study, 39% of the patients had used psilocybin 10 or more times, with 22% reporting “hundreds” of experiences with psychedelics or entactogens (Anderson et al., 2020). Quality of the eight qualitative studies was high; one study reported no AEs (Noorani et al., 2018).

Physiological data

Elevated blood pressure (BP) in the psilocybin group was the most common physiological effect, which was significant in three studies that performed tests of significance (Bogenschutz et al., 2015; Ross et al., 2016; Schindler et al., 2020). Increased HR in the psilocybin group was significant in one study but not in two others. None of the elevations in vital signs required medical intervention. Self-limiting severe hypertension (systolic BP > 180 or diastolic BP > 110 mmHg) occurred in 4 of 18 patients in one study, which seemed related to severe anxiety and resolved upon reassurance by therapists (Anderson et al., 2020). In another study, BP exceeded protocol criteria (diastolic BP > 100 mmHg) in one patient, which resolved without intervention (Davis et al., 2020).

Adverse events

Depressive disorders

A small open-label study on psilocybin treatment for TRD reported anxiety, confusion, and nausea as the most prevalent acute AEs (Carhart-Harris et al., 2016). Anxiety occurred mostly prior to sessions and at onset, whereas confusion occurred during the peak of the session. Severity was mild to moderate, except for one case of severe anxiety during a high-dose session. Both MDD RCTs reported headaches (Carhart-Harris et al., 2021; Davis et al., 2020). In one RCT, patients reported moderate to strong emotional and psychological AEs, with patients endorsing statements such as “I felt like crying,” sadness, emotional and/or physical suffering, feelings of grief, and feelings of isolation (Davis et al., 2020). One participant experienced confusing traumatic memories, in which a parent held a pillow over his face, which worsened his depression for several weeks post-treatment (Carhart-Harris et al., 2018; Watts et al., 2017); another patient became incommunicative during the psilocybin session (Carhart-Harris et al., 2018). Two qualitative studies illustrated experiences from a patient perspective:

I worried that I let [the music] shape this sort of melancholy. There was resistance, massively, to everything, every sort of sensory input, I had a fearful response. I was afraid to open my eyes, I was afraid to do anything, I was afraid that this sort of music was the last thing I’d ever hear. (Kaelen et al., 2018)

There was a lot of sadness, really really deep sadness: the loss the grief, it was love and sadness together, and letting go, I could feel the grief and then let it go because holding onto it was hurting me, holding me back. It was a process of unblocking. (Watts et al., 2017)

Post-treatment headache was the only reported late AE; the open-label study only reported headaches in the high-dose (25 mg) group. In one study, 67% rated one of the psilocybin sessions among their top five most psychologically challenging experiences (with 30% rating it the single most psychologically challenging experience (Davis et al., 2020); this was not asked in the other studies (Carhart-Harris et al., 2021).

End-of-life anxiety

Two EOLA studies reported acute and transient AEs in the psilocybin groups, including physical discomfort (headache, nausea) and psychological discomfort (anxiety, transient thought disorders, and suicidal ideation) (Griffiths et al., 2016; Ross et al., 2016). One study did not report on AEs (Grob et al., 2011). Only one study reported headache as a late AE (Griffiths et al., 2016). One patient committed suicide 11 days after a low (placebo like) dose (1 mg) psilocybin session. The patient, reporting feeling bored, wanted to leave this session early, and was subsequently discontinued from the study. The authors concluded that it was an SAE not related to research procedures or to psilocybin (Griffiths et al., 2016).

Three qualitative studies described patient experiences:

It really hit me very strong. And um, it was terrifying. It was just terrifying. It was, um, I was completely disoriented . . . I was really, maybe, in the hold of a ship at sea. Rocking. Absolutely nothing, nothing to anchor myself to, nothing, no point of reference, nothing, just lost in space, just crazy, and I was so scared. And then I remembered that [the therapists] were right there and suddenly realized why it was so important that I get to know them and they to get to know me. And reached out my hand and just said “I’m so scared.” And I think it was [the therapist] who took my hand . . . and said “It’s all right. Just go with it. Go with it.” And um, and I did. (Belser et al., 2017)

Substance use disorders

Two open-label studies investigated psilocybin as a treatment for SUD (Bogenschutz et al., 2015; Johnson et al., 2014). One study reported strong (5 of 15 patients) to extreme (1 of 15 patients) anxiety and feeling trapped as acute AEs, and headache as a late AE (2014). The other study reported few AEs (Bogenschutz et al., 2015), although qualitative studies illustrated how frightening the experience was for some:

That whole experience just felt for me like, a fever nightmare. (Nielson et al., 2018)

In the second session, [she] received a higher dose of medication and experienced an amplification of thought moving her into a confused and chaotic state. Underneath the chaotic thinking, she identified a deep well of overwhelming sadness. She was able to eventually surrender control over her thoughts and entered into a state of peacefulness, until her thoughts quieted completely. (Bogenschutz et al., 2018)

Migraines

Most reported acute AEs in the migraine RCT were nausea, anxiety and lightheadedness; headache and migraine attacks were the only late AEs (Schindler et al., 2020). All were more prevalent in the psilocybin group.

Demoralization

The most commonly reported acute AEs in the open-label study on demoralization in gay men were anxiety, nausea, and headache (Anderson et al., 2020). Most of these patients (14 of 18) had transient AEs; nearly half of them reported these AEs as moderate to severe (7 of 18) AEs. Two unexpected late AEs occurred: one patient experienced a post-traumatic flashback, leaving him unable to work for 2 days. Another participant experienced a severe exacerbation of anxiety, followed by a relapse in methamphetamine use, leading to study withdrawal (Anderson et al., 2020).

Ayahuasca

We included three studies, describing 48 patients in total: one RCT in patients with a social anxiety disorder (SAD) (Dos Santos et al., 2021), one RCT in TRD patients (Palhano-Fontes et al., 2019), and one open-label study in MDD patients (Osorio et al., 2015; Sanches et al., 2016), all involving a single administration of ayahuasca.

Quality appraisal

None of the studies systematically assessed AEs, and only one reported acute AEs spontaneously reported or observed by therapists (Sanches et al., 2016). At study commencement, all participants were naïve to ayahuasca and most were naïve to other serotonergic psychedelics as well.

Physiological data

No significant HR or blood pressure elevations were observed (Dos Santos et al., 2021; Osorio et al., 2015; Sanches et al., 2016); one study did not report on physiological parameters (Palhano-Fontes et al., 2019).

Adverse events

Nausea, vomiting, diarrhea, and anxiety were more common acute AEs in the ayahuasca group, whereas anxiety and headache were more prevalent in the control group in one study (Palhano-Fontes et al., 2019). In the TRD study, four patients remained hospitalized for seven days, due to a ‘more delicate condition’ (Palhano-Fontes et al., 2019). In the SAD study, one participant experienced an intense episode of transient fear of dying and/or going crazy, distress, and dissociation, which resolved after reassurance by therapists (Dos Santos et al., 2021). In this study, no AEs were reported in the control group. The open-label study only reported vomiting as an acute AE (Osorio et al., 2015; Sanches et al., 2016). No studies reported on late AEs.

LSD

Four studies were included, describing 27 patients in total: one RCT with a nested qualitative study describing EOLA treatment (Gasser et al., 2014, 2015); one observational study on compassionate use of LSD and/or MDMA (Schmid et al., 2020) in group therapy with patients, mostly with PTSD and/or MDD; and one case report on LSD treatment for a complex personality disorder (Müller et al., 2020). Active doses of 100–200 μg were used.

Quality appraisal

AEs were collected through spontaneous reporting by patient and therapist observations. 39% of participants in the compassionate group therapy (Schmid et al., 2020) had prior drug (including cannabis) experience; only one participant (8%) in the RCT did (Gasser et al., 2014). Quality of the qualitative study was medium/high. AE severity was reported only in the RCT (Gasser et al., 2014).

Physiological data

No significant between-group differences were observed on HR or blood pressure (Gasser et al., 2014).

Adverse events

Acute, transient AEs included illusions, feeling cold, feeling abnormal, and anxiety; all but moderate-to-severe anxiety were more prevalent in the high (active) dose groups. Mild-to-moderate emotional distress was of equal severity in the low- and high-dose group. Illusions, feeling abnormal, and feeling cold were also late AEs for one/two patients in the high-dose group. Severity was mostly mild to moderate. Examples of emotional distress and anxiety, as well as later resolution, were described by one patient as follows:

“The first trip was a panic trip. With almost pure fear of death. It was agony . . . Really, I had the feeling ‘that I am dying’. Yes, it was just really black, the black side. I was afraid, shaking. . . . It was total exhaustion, not seeing an exit, no escape. It seemed to me like an endless marathon . . . that was a big part of the trip until it finally led to relaxation . . . During the second trip, the dark side also showed up at the beginning, but for a rather short time. I was a little tensed, sweating, but not for so long and suddenly a phase of relaxation came. Completely detached. It became bright. Everything was light. It became a pleasant feeling, a warm feeling. No pain. Almost a little floating, clear, being carried and together with the music. . . It was really gorgeous. . . . The key experience is when you get from dark to light, from tension to total relaxation” (Gasser et al., 2015)

Discussion

This is the first overview of AEs associated with the clinical use of MDMA and classic serotonergic psychedelics. Using a mixed-methods sequential explanatory design, we combined quantitative measures with qualitative descriptions of patient experiences. Based on this review, MDMA and psychedelics appear to be well tolerated in the treatment of a range of different (mental) disorders. However, caution is warranted, since this conclusion is based on a limited number of relatively small studies with a variety of low- and high-quality designs, using different (and inconsistent) AE assessment procedures, different types of psychedelics used in different doses, in patients with different disorders with relatively high rates of prior experience with psychedelic drug use.

Acute AEs

The most common acute physical AEs in the MDMA groups were fatigue, lack of appetite, feeling cold, thirst, jaw clenching, and perspiration; anxiety and difficulty in concentrating were the most common acute psychological AEs. The most common acute AEs across psilocybin studies were moderate to severe anxiety, headache, and nausea. Acute psychological AEs for psilocybin and LSD (e.g., paranoid thoughts, feeling trapped, illusions, feeling abnormal, psychological discomfort) mostly resolved during sessions, but were sometimes severe. In one ayahuasca study (Dos Santos et al., 2021), a participant temporarily dissociated, fearing he/she was dying or going crazy. Therapist interventions helped resolve this episode by calming this patient down, emphasizing the importance of trained, skilled, and trusting therapists (Mithoefer, et al., 2013a; Phelps, 2017, 2019). Therapist training is also important to familiarize therapists with the contextuality of some AEs, and of effective interventions to prevent or manage them.

Multiple qualitative studies describing occasionally terrifying, frightening, and confusing experiences. In many cases, patients evaluated these experiences retrospectively as therapeutic and beneficial. Nausea and vomiting were common acute AEs in ayahuasca studies. Except for a single case of premature ventricular contractions in one MDMA-treated patient with PTSD, no drug-related AEs and SAEs required medical intervention.

Many of the physiological AEs described in this review are well known from the literature on healthy subjects: for example, nausea, and vomiting with ayahuasca (Dos Santos et al., 2016; Durante et al., 2021; Guimarães dos Santos, 2013; Hamill et al., 2019), headache after psilocybin use (Johnson et al., 2012), and acute, transient elevated vital signs, jaw clenching, perspiration, and lack of appetite with MDMA (Dumont and Verkes, 2006; Liechti et al., 2000; Vizeli and Liechti, 2017; Vizeli et al., 2019; Vollenweider et al., 1998). Across studies and disorders, acute psychological AEs included anxiety, transient paranoia, confusion, despair, and psychological distress were commonly reported, ranging from relatively mild to severe in some cases; most of the AEs resolved during the session.

Late AEs

The most common late AEs across MDMA studies were fatigue, headache, anxiety, difficulty concentrating, and low mood. A minority of patients reported anxiety, depression, panic, and suicidal ideation in the days and weeks following treatment sessions. Low mood was a late AE present in both MDMA and placebo groups. Low mood is also commonly reported, 3–4 days after MDMA use, in studies with healthy subjects and by recreational “ecstasy” users (Baggott et al., 2016; Dolder et al., 2018; Kirkpatrick and De Wit, 2014; Liechti et al., 2000, 2001; Vollenweider et al., 1998). However, some studies do not report post-acute decreases in mood (Borissova et al., 2021), including the open-label study on MDMA in patients with an AUD (Sessa et al., 2021a). Patients in control groups may have experienced low mood due to a lack of symptom relief and related disappointment; anxiety in the MDMA groups may be related to re-experiencing traumatic memories and painful emotions. Future research should more accurately qualify anxious states and low mood in the period following MDMA treatment, and to assess whether these are a time-limited element of the therapeutic process or a subacute adverse effect.

One ayahuasca study reported hospitalizing four patients for a week, without further commenting how their ‘delicated condition’ related to the (adverse) effects of ayahuasca (Palhano-Fontes et al., 2019). Two studies (Anderson et al., 2020; Watts et al., 2017) reported unexpected adverse reactions (post-traumatic flashback; severe exacerbation of anxiety and drug use relapse; exacerbation of depression) in three patients that occurred several weeks post-psilocybin administration; patients’ conditions deteriorated and required further therapeutic intervention; one patient withdrew from the study. The authors note that these reactions and the high incidence of anxiety, weeks after the session, may be related to the medical complexity and high psychiatric comorbidity of this population (Anderson et al., 2020). Despite the clinical severity of these complaints, they were not reported as SAEs, which carries the risk of underreporting potential negative consequences. On the other hand, the longer the time after treatment sessions, the less likely it is that events are still related to a specific treatment. Since we currently cannot predict individual responses and adverse reactions that may develop in the weeks or months following treatment, it is imperative that researchers and clinicians perform longer term follow-ups and report any untoward reactions.

Recently, several cases were reported in which study participants in an MDMA trial stated that they became suicidal after the trial was finished, including one case in which a participant remained attached to a therapist, and was subsequently abused (Lindsay, 2022; Rosin, 2022). Unfortunately, such very serious transgressions occur in our field, and need to be vigorously addressed if and when they occur (Gutheil and Gabbard, 1993). It might well be that the field of psychedelic treatments is especially at risk, as patients under influence are vulnerable, suggestible and highly dependent on the therapist(s) (Hartogsohn, 2018). Furthermore, therapists may have developed their own personal approaches to therapy that not always adhere to professional codes of conduct (Gutheil and Gabbard, 1998). Apart from that, this also illustrates the importance of thoroughly and systematically monitoring adverse effects.

Suicidality

We found one study in which a suicide, 11 days after a session with a placebo-like dose (1 mg) of psilocybin, was reported as unrelated to study participation (Griffiths et al., 2016). Not included in our review were the results of a recently finalized RCT, comparing a very low (1 mg) to a medium (10 mg) and a high dose (25 mg) of psilocybin in patients with TRD (n = 233). Overall AE rates seem roughly comparable to those summarized in this review. A preliminary (non-peer-reviewed) publication reported suicidal ideation, suicidal behavior, and/or intentional self-injury—in 7 of the 79 patients (8.8%) in the 25 mg group, in 6 of the 75 patients (8%) in the 10 mg group, and in 4 of the 79 patients (5%) in the 1 mg group; 14 instances in 9 patients were reported as SAEs (Compass Pathways, 2021). These numbers suggest higher suicidality rates in active dose groups (10 and 25 mg), but further interpretation can only be made once peer-reviewed results have been published. Suicidal ideation and behavior were reported in three MDMA studies, although incidence was low, rates were lower or similar in MDMA compared to controls, and SAEs were only reported in the control groups. Nevertheless, scrutinous assessment of treatment-related consequences is warranted. Being randomized to the inactive group, combined with despair and (overly) high expectations in patients with severe disorders may contribute to feelings of hopelessness, and subsequent suicidal ideation and/or behavior. That said, it is important to remember that mental disorders increase the risk of suicidal thoughts in general (Andersson et al., 2022; Nock et al., 2009). Future studies need to carefully asses suicidal ideation, and to explore ways to best support all patients in the long run.

Interpreting adversity

Psychological or emotional “AEs” can be ambiguous, subjective, context dependent, and subject to interpretation. The qualitative studies in this review illustrated how, in hindsight, patients considered working through difficult feelings and emotions as therapeutically beneficial (Barrett et al., 2016; Roseman et al., 2019). This is corroborated by (survey) studies showing that users frequently transform challenging experiences into personally, morally, or spiritually meaningful ones (Barrett et al., 2016; Carbonaro et al., 2016; Gashi et al., 2021; Roseman et al., 2019). The shift in terminology—using “challenging experiences” instead of “bad trips” (AEs)—indicates an increasing acceptance that difficult experiences are inherent to psychedelic treatments (Johnson et al., 2008), and that they can be both distressing and therapeutically beneficial. More narrative evidence as well as the systematic use of specialized questionnaires (e.g., CEQ) can help interpret such experiences and their impact on treatment outcomes. This raises some important issues.

First, since the effects of psychedelics are thought to be influenced substantially by contextual factors, this may extend to adverse effects as well. Treatment designs that reduce or minimize positive contextual components (e.g., time spent preparing patients, number of therapists, strength of the therapeutic relationship, time spent providing aftercare and integration) may increase the incidence of AEs (Erritzoe and Richards, 2017; Oram, 2012, 2016). Second, some aspects of the definitions of “AE” are ill-suited in the context of psychedelic treatment. By definition, AEs are undesired and something to be avoided. Categorizing challenging but potentially therapeutically beneficial experiences as AEs can thus be counterproductive. For instance, framing “anxiety” as adversity may prompt patients to try and avoid difficult feelings and thoughts rather than engaging with them as part of a therapeutic process. Disentangling adverse elements that ought not be part of the therapeutic process (e.g., physical AEs such as headaches and nausea, or exacerbation of distressing symptoms such as suicidal thoughts and dissociation) from psychological experiences that are part of the therapeutic process is crucial to improve our understanding of the mechanisms of action of psychedelic treatments. To illustrate the complexity of this challenge: vomiting, the most commonly reported AE in ayahuasca studies, is often considered a therapeutic element both in traditional indigenous and in “neoshamanic” ayahuasca practices (Fotiou and Gearin, 2019). In the case of anxiety, the severity and the length of the difficult episode may also predict suboptimal or even negative clinical outcomes (Carbonaro et al., 2016; Roseman et al., 2018). Careful screening, preparation, session monitoring, psychological support, and post-session integration are crucial prerequisites to maintain a low incidence of AEs and to maximize therapeutic efficacy (Johnson et al., 2008; Strassman, 1984). In the absence thereof, young people, and those high in the personality traits of neuroticism emotional lability, and/or low in absorption, openness to experience, and acceptance, as well as those in a apprehensive, preoccupied or confused mental state may be more prone to challenging experiences, which can worsen their pre-existing problems (Aday et al., 2021; Barrett et al., 2017; Studerus et al., 2012). Finally, recent articles have described how psychedelics may alter metaphysical or ontological beliefs (Nayak and Griffiths, 2022; Timmermann et al., 2021). Such experiences may be highly meaningful but can also induce an “ontological shock,” particularly when conflicting with prior held religious, personal, or spiritual beliefs (Breeksema and Van Elk, 2021). This introduces ethical and metaphysical challenges, and a deep responsibility for therapists to mitigate potential adversity when patients are confronted with profound metaphysical experiences (Timmermann et al., 2022).

Limitations

The current review has both strengths and limitations. An important strength is the systematic collection of studies and the combination of quantitative and qualitative studies. However, this review also has some important limitations. A first limitation is that, included studies were mostly small, varied in study design, (dose of) psychedelic drug, and demographics, making comparisons difficult and performing a meta-analysis impossible. Second, only few studies assessed AEs consistently or systematically. Unfortunately, this is a general problem in psychopharmacology (Coates et al., 2018) and not unique for psychedelic drug studies. Instead, most studies relied on spontaneously reported and/or therapist-observed AEs. Spontaneous reporting depends on individual and contextual factors enabling patients to recall or discuss negative experiences. Therapist observation is similarly ambiguous, as observations may be colored by their subjective perception and interpretation. Five pilot studies did not report any AEs (Bouso et al., 2008; Grob et al., 2011; Moreno et al., 2006; Searchfield et al., 2020; Sessa, et al., 2021b). While it is possible—although unlikely—that no AE occurred, not reporting any AE is reminiscent of important omissions in early psychedelic research, when researchers often did not fully disclose the frequency and severity of negative reactions (Rucker et al., 2018). Third, there are not enough placebo-controlled trials to warrant strong conclusions about the causal relation between the psychedelic treatment and AEs; effectively blinding study participants to their treatment condition is notably hard in psychedelic drug studies (Aday et al., 2022; Muthukumaraswamy et al., 2021), particularly given the high percentages of patients who have had previous experience with a psychedelic. Also, due to the small sample sizes, studies may have missed AEs that occur less frequently, hindering extrapolation as some AEs may only become apparent in larger populations. All studies in our review refer to a total of 598 unique patients—of whom 521 received an active dose. Fourth, severity and onset timing of AEs (particularly for late AEs) was reported in only a few studies, hampering adequate interpretation of reported AEs. Next, most trials had strict criteria for participation, with extensive screening and excluding patients with almost any comorbid mental disorder. This includes a personal (or first-degree family) history of psychotic, bipolar, dissociative identity, eating, and SUDs, and active suicidal ideation, all of which are frequently found in patients with more severe mental disorders. Sixth, many studies included patients with previous experience with the experimental drug; up to 56% for MDMA and roughly half of all psilocybin study participants had prior experience(s) with psilocybin or other serotonergic psychedelics. Intensity of drug effects and experienced difficulties appear to be lower and less severe in those with past experience with these drugs (Aday et al., 2021), limiting generalizability to broader patient populations. Moreover, minorities are underrepresented in clinical research with psychedelics (Michaels et al., 2018). It is, therefore, possible that psychedelic-naïve patients with a different socio-cultural background and more severe forms of psychopathology may have more intense experiences, or experience more difficulties coping with potentially disruptive events (Griffiths et al., 2006). That said, the only study in our review that compared lifetime psychedelic use and degree of experienced challenge within the session (as measured with the CEQ) found no significant correlations, suggesting that previous experience does not necessarily predict the occurrence of AEs in clinical populations or contexts (Anderson et al., 2020). Seventh, while qualitative studies helped contextualize some of the acute challenging experiences, these were not always specifically designed to explore such effects. Furthermore, qualitative studies are unevenly divided and relatively scarce; there were none on ayahuasca, only one on MDMA and LSD, and eight on clinical psilocybin studies. Finally, any conclusion on the frequency and interpretation of AEs should consider the target population (set) and the context of the treatment (setting). More information should be given about “set and setting” when reporting and interpreting the results of clinical trials with psychedelics.

Conclusions

This first review of acute and late AEs in the treatment with serotonergic psychedelics and MDMA, combining quantitative and qualitative studies, shows that all compounds acutely induce transient headaches, nausea, and anxiety. Anxiety covers a wide range of psychologically challenging events that may also be beneficial for the therapeutic process. Disentangling truly AEs from potentially beneficial effects is complex but crucial to improve our understanding of psychedelic treatments. SAEs seem to be largely absent based on the included studies, as were lasting physiological side effects. Suicidal ideation and suicidal behavior were rare but did occur; these constitute a serious psychiatric emergency, which emphasizes the importance to investigate which patients are most at risk and how to best reduce the likelihood of their occurrence. Overall, AEs were inconsistently defined and inadequately assessed. Future studies should describe timing and severity of effects more extensively, for example, using scales such as the CEQ. Many studies included patients with prior psychedelic experience, which may bias results and limit generalizability. Qualitative research can also add nuance by detailing and understanding the meaning of challenging experiences. Full transparency about AEs is a responsibility of clinicians, particularly in a nascent field fueled by the enthusiasm of pioneering researchers. Understanding the full spectrum of unpleasant, potentially harmful, and transformative treatment-related events is crucial to inform future therapists who may otherwise be sub-optimally prepared to handle challenging and potentially destabilizing patient experiences, particularly in larger groups of patients with more complex and potentially comorbid conditions.

Introduction

Clinical research exploring psychedelic compounds has shown promising outcomes for various mental health conditions. However, the studies typically involved small and select groups of participants. With the first larger clinical trials concluding and new applications being explored, some of these compounds are approaching approval for medical use.

The non-medical use of psychedelics can lead to immediate negative reactions such as anxiety, panic, extreme sadness, suspicion, and unsafe behaviors. Such events might also contribute to lasting psychological issues. While these issues have been mainly linked to non-medical use, and small clinical trials have shown safety, using these substances in larger, more diverse patient populations with multiple health problems could lead to unexpected negative effects. Additionally, unwanted drug reactions can cause patients to stop treatment. Difficult experiences that are not processed well could also lead to ongoing negative outcomes. Therefore, it is important to identify all possible adverse reactions to psychedelic drugs, especially in vulnerable patients with mental disorders that are hard to treat.

Information on this topic has not been systematically reviewed since 1984. An earlier review from 1960, based on responses from clinicians treating approximately 5,000 patients (mostly with LSD), concluded that negative events were rare, with some exceptions mainly in patients with schizophrenia, and that the drugs were generally safe when administered carefully. A 1984 review of negative reactions to psychedelics in different settings found that these events occurred on a spectrum, from immediate, temporary panic during administration, to short-term psychosis lasting several days, and even recurring flashbacks or chronic psychotic conditions. Flashbacks were later included in the diagnosis "hallucinogen persisting perception disorder" (HPPD), which appears in one major diagnostic manual but not another. HPPD is reported occasionally, mostly after frequent non-medical use of certain psychedelics or MDMA, and may be linked to existing mental health conditions. However, flashbacks are often described as mild, neutral, or even pleasant. A more recent review explored whether commonly discussed harms were based on personal stories or scientific evidence.

Assessing negative events is challenging for several reasons. These events are not always defined beforehand, the range of possible reactions can be wide, reporting can be inconsistent, and terms are often used interchangeably, such as "side effects," "toxic effects," "adverse effects," and "harms." This is further complicated by the highly varied and situation-dependent personal experiences that psychedelics can cause.

As more patients with mental health conditions may soon be treated with psychedelics, a complete understanding of both the helpful effects and any negative events is necessary for a balanced assessment of benefits and risks, and to determine which patients are most likely to benefit or not. The current paper aims to systematically review all negative events occurring during or after psychedelic treatments with classic hallucinogens (such as psilocybin, LSD, ayahuasca) and MDMA in patients. While MDMA and classic hallucinogens have pharmacological differences, they are often grouped together as "psychedelics" and used in similar therapeutic settings. Negative events associated with ketamine and ibogaine have been reviewed elsewhere.

Methods

Given the relative newness of this field, this review takes an exploratory approach rather than trying to confirm specific hypotheses. The review was not registered in advance. This systematic review and data summary follows a step-by-step design, using established guidelines to gather, organize, and summarize findings from both quantitative and qualitative studies. The term "adverse event" (AE) refers to any unfavorable or harmful event that first occurs during or after the administration of a classic psychedelic or MDMA, regardless of its connection to the treatment. In this review, acute AEs refer to those occurring during or on the day of a psychedelic or MDMA session, while late AEs refer to those appearing after the session day. The term "treatment-emergent AEs" (TEAEs) is sometimes used to describe an AE that either begins during treatment or an existing one that worsens. Since there is no clear difference between these terms, this review only distinguishes between acute and late AEs.

Selection criteria

All quantitative and qualitative clinical studies describing the treatment of patients with a mental disorder using a classic serotonergic psychedelic (e.g., psilocybin, LSD, mescaline, ayahuasca) or an entactogenic drug (e.g., MDMA, MDA), including open-label studies, case reports or series, and randomized controlled trials (RCTs) published since 2000, were included. Systematic reviews, surveys, secondary analyses, and studies involving healthy volunteers were excluded. The goal was to assess possible negative reactions in patients diagnosed with a mental disorder who were treated with any of these compounds. Therefore, no specific exclusion criteria were set beforehand, allowing for the broadest possible inclusion to avoid missing any potential unfavorable outcome.

Search methods

A systematic search of PubMed/Medline, EMBASE, and PSYCinfo databases was conducted on July 28, 2021. The search combined terms for psychedelic compounds (e.g., psilocybin, lysergic acid diethylamide, MDMA) and study types (e.g., clinical trial, qualitative), excluding animal studies. ClinicalTrials.gov was also searched to cross-check corresponding publications. Systematic searches were supplemented by manual searching and checking reference lists.

Quality assessment

A formal quality assessment was not considered strictly necessary because even high-quality studies with low bias might not report AEs well. Instead, specific quality criteria relevant to the review's aim were assessed: the methods used to monitor or report AEs, study inclusion and exclusion criteria, and the percentage of participants with previous experience with the drug. A specific checklist was used to assess the methodological rigor of qualitative studies.

Data extraction and synthesis

All descriptive and quantitative information on acute and late AEs, the timing of assessment, and serious AEs (SAEs) were extracted from the main text and supplementary files of the publications. The prevalence of AEs was recalculated as a percentage of total participants. Qualitative articles were carefully reviewed for themes and descriptions related to AEs. These were then used to add to, illustrate, and provide context for findings from quantitative studies.

Results

A total of 5,640 articles were found. After removing duplicates, 3,709 titles and abstracts were screened. Following screening, the full texts of 61 articles were assessed for eligibility, and 44 articles were included. Additionally, 596 registered trials were found, of which 24 completed trials were included.

The included studies were published between 2006 and 2021 and reported on a total of 598 unique participants, with 521 receiving an active dose. Sample sizes ranged from 1 to 105 participants, with average ages between 25 and 59 years. Multiple publications referring to a single study were combined. All qualitative studies described smaller groups taken from quantitative studies included in this review. There was significant variation in study design, substances, dosages, and disorders. Four quantitative and two qualitative studies did not report on AEs.

MDMA

Sixteen studies were selected, including one qualitative study, involving a total of 266 patients treated with MDMA. Ten publications described MDMA-assisted treatment for Post-Traumatic Stress Disorder (PTSD), including seven randomized controlled trials (RCTs), one open-label trial, one case series, and one qualitative study. Other RCTs reported on the treatment of end-of-life anxiety (EOLA), social anxiety in adults with autism, and tinnitus. One open-label study focused on alcohol use disorder (AUD). Active doses ranged from 50 to 125 mg, sometimes followed by an optional supplemental half-dose.

Quality appraisal. Of the MDMA studies, eleven reported AEs only when spontaneously mentioned by participants; one study did not report on AEs. Two RCTs systematically assessed AEs, and another study used a specific scale to report AE severity, the only one to do so. Six studies did not report whether participants had previously used MDMA. Prior MDMA exposure in the other nine quantitative studies varied from none to 56%. The qualitative study was of medium to high quality but did not report on AEs.

Physiological effects. Nine studies measured blood pressure, body temperature, and heart rate. All reported mild, temporary, and statistically significant increases during the MDMA session. None required medical intervention. One MDMA-related serious adverse event (SAE) occurred when a participant experienced an increase in irregular heartbeats, which were also present at baseline. This required one night of hospital monitoring and resolved on its own.

Adverse events. For PTSD, the most common immediate physical AEs included jaw clenching, muscle tightness, headaches, nausea, fatigue, and lack of appetite. Other immediate AEs were feeling cold, thirst, dizziness, sweating, restlessness, and body pains. Anxiety was the most common psychological AE. Most immediate AEs occurred more often in the MDMA groups than in placebo groups, though some exceptions existed. Immediate AEs were mild to moderate in severity in the single study that reported severity. No clear relationship between dose and AE was found in the two PTSD studies that assessed different active MDMA doses. SAEs were reported in two studies: in one, two patients reported suicidal behavior or thoughts leading to self-hospitalization, but these occurred only in the control group. In another study, three of four SAEs were considered unrelated to the treatment. One patient developed an exacerbation of pre-existing irregular heartbeats, which resolved without evidence of heart disease and was considered potentially related.

A case study illustrated how a participant experienced these immediate adverse reactions and how they related to the therapeutic process. The participant experienced strong emotional reactions and physical sensations while revisiting traumatic memories. Afterward, the participant reflected on the experience as needing to "work through the darkness."

Common late AEs included fatigue, lack of appetite, low mood, insomnia, increased need for sleep, irritability, headache, difficulty concentrating, and anxiety. In one study, rates of fatigue, insomnia, and headaches were somewhat higher in the control group. A minority of patients in the MDMA group reported complaints up to two months after the final session, including anxiety, depressed or low mood, and panic attacks, with higher rates in the MDMA group compared to the control group.

For end-of-life anxiety, a crossover RCT reported jaw clenching, thirst, dry mouth, sweating, and headache as immediate AEs. Fatigue, increased need for sleep, insomnia, anxiety, jaw clenching, and low mood were the most common late AEs. Except for low mood, all AEs occurred more often in the MDMA group. Anxiety, depressed mood, and dissociation were reported only in the MDMA group; insomnia was reported in both groups.

For social anxiety, an RCT in adults with autism reported higher rates of immediate AEs in the MDMA group than in the placebo group, including anxiety, difficulty concentrating, fatigue, headache, and sensitivity to cold. Most prevalent late AEs included fatigue, headache, difficulty concentrating, low mood, and the need for more sleep; all late AEs except the latter two were more common in the MDMA group. Two of eight patients in the MDMA group and one of four patients in the placebo group reported depressed mood, suicidal thoughts, and panic attacks in the period following the first drug session.

Studies on tinnitus and AUD treatment did not report on AEs.

Psilocybin

Twenty articles were included, describing psilocybin treatment in 257 patients: six crossover RCTs, five articles describing four open-label studies, one single-group, pseudo-randomized dose-escalation study, and eight qualitative studies. Three RCTs focused on EOLA, two RCTs on major depressive disorder (MDD), one RCT on migraine headaches, and one dose-escalation study on obsessive-compulsive disorder. Open-label studies investigated the treatment of treatment-resistant depression (TRD), AUD, tobacco smoking, and demoralization in older AIDS-survivors. Active dosages varied from 10 to 30 mg or 0.1 to 0.4 mg/kg of psilocybin per session. Only two of six studies that administered different doses reported on adverse effects per dosage; one study described the timing of AEs per session. Qualitative studies described treatment experiences among patients with EOLA, TRD, and substance use disorder (SUD).

Quality appraisal. Two studies did not report on AEs. Most other studies relied on spontaneous reporting by patients and observations by therapists. Some also used post-treatment interviews, headache ratings, and structured questionnaires. Three studies reported on AE severity. In studies that assessed dose-effect relationships, one found a higher prevalence of some psychological AEs in the moderate-dose (20 mg) than in the high-dose (30 mg) session, and another reported a higher prevalence of post-treatment headaches in the higher dose group. One study reported a much higher number of distinct AEs than all other studies.

With one exception, all quantitative studies reported on prior psychedelic use, which varied from 20% to 100%, averaging 50% of patients having used psilocybin or other psychedelics before the study. One study reported a group average of 0.8 lifetime experiences with psychedelics; in another study, 39% of patients had used psilocybin 10 or more times, with 22% reporting "hundreds" of experiences. The quality of the eight qualitative studies was high; one study reported no AEs.

Physiological data. Elevated blood pressure in the psilocybin group was the most common physiological effect, significant in three studies that performed significance tests. Increased heart rate in the psilocybin group was significant in one study but not in two others. None of the vital sign elevations required medical intervention. Self-limiting severe high blood pressure occurred in 4 of 18 patients in one study, which seemed related to severe anxiety and resolved with reassurance from therapists. In another study, blood pressure exceeded protocol criteria in one patient and resolved without intervention.

Adverse events. For depressive disorders, a small open-label study on psilocybin treatment for TRD reported anxiety, confusion, and nausea as the most prevalent immediate AEs. Anxiety mostly occurred before sessions and at onset, while confusion occurred during the session's peak. Severity was mild to moderate, except for one case of severe anxiety during a high-dose session. Both MDD RCTs reported headaches. In one RCT, patients reported moderate to strong emotional and psychological AEs, endorsing statements such as "I felt like crying," sadness, emotional or physical suffering, feelings of grief, and feelings of isolation. One participant experienced confusing traumatic memories that worsened depression for several weeks post-treatment; another patient became incommunicative during the psilocybin session. Qualitative studies illustrated patient experiences of fear, disorientation, deep sadness, and the process of "unblocking."

Post-treatment headache was the only reported late AE; the open-label study reported headaches only in the high-dose (25 mg) group. In one study, 67% of patients rated one of the psilocybin sessions among their top five most psychologically challenging experiences.

For end-of-life anxiety, two studies reported immediate and temporary AEs in the psilocybin groups, including physical discomfort (headache, nausea) and psychological discomfort (anxiety, temporary thought disorders, and suicidal thoughts). One study did not report on AEs. Only one study reported headache as a late AE. One patient committed suicide 11 days after a low (placebo-like) dose (1 mg) psilocybin session. The patient wanted to leave this session early due to boredom and was subsequently removed from the study. The authors concluded it was an SAE unrelated to research procedures or psilocybin. Qualitative studies described patient experiences of terror, disorientation, feeling lost, and eventually finding reassurance and surrendering to the experience.

For substance use disorders, two open-label studies investigated psilocybin. One study reported strong (5 of 15 patients) to extreme (1 of 15 patients) anxiety and feeling trapped as immediate AEs, and headache as a late AE. The other study reported few AEs, although qualitative studies illustrated how frightening the experience was for some. One participant described the experience as a "fever nightmare," while another experienced amplified, chaotic thoughts that eventually led to a state of peacefulness.

For migraines, most reported immediate AEs in the RCT were nausea, anxiety, and lightheadedness; headache and migraine attacks were the only late AEs. All were more prevalent in the psilocybin group.

For demoralization, the most commonly reported immediate AEs in the open-label study in gay men were anxiety, nausea, and headache. Most of these patients (14 of 18) had temporary AEs; nearly half reported these as moderate to severe. Two unexpected late AEs occurred: one patient experienced a post-traumatic flashback, preventing them from working for two days. Another participant experienced a severe worsening of anxiety, followed by a relapse in methamphetamine use, leading to study withdrawal.

Ayahuasca

Three studies were included, describing 48 patients in total: one RCT in patients with social anxiety disorder (SAD), one RCT in TRD patients, and one open-label study in MDD patients, all involving a single administration of ayahuasca.

Quality appraisal. None of the studies systematically assessed AEs, and only one reported immediate AEs that were spontaneously mentioned or observed by therapists. At the start of the study, all participants were new to ayahuasca, and most were new to other serotonergic psychedelics as well.

Physiological data. No significant increases in heart rate or blood pressure were observed; one study did not report on physiological parameters.

Adverse events. Nausea, vomiting, diarrhea, and anxiety were more common immediate AEs in the ayahuasca group. Anxiety and headache were more prevalent in the control group in one study. In the TRD study, four patients remained hospitalized for seven days due to a "more delicate condition." In the SAD study, one participant experienced an intense episode of temporary fear of dying or losing control, distress, and dissociation, which resolved after reassurance from therapists. In this study, no AEs were reported in the control group. The open-label study only reported vomiting as an immediate AE. No studies reported on late AEs.

LSD

Four studies were included, describing 27 patients in total: one RCT with a nested qualitative study describing EOLA treatment; one observational study on compassionate use of LSD or MDMA in group therapy with patients, mostly with PTSD or MDD; and one case report on LSD treatment for a complex personality disorder. Active doses of 100–200 μg were used.

Quality appraisal. AEs were collected through spontaneous reporting by patients and observations by therapists. In the compassionate group therapy, 39% of participants had prior drug experience; only one participant (8%) in the RCT did. The quality of the qualitative study was medium to high. AE severity was reported only in the RCT.

Physiological data. No significant differences between groups were observed in heart rate or blood pressure.

Adverse events. Immediate, temporary AEs included illusions, feeling cold, feeling abnormal, and anxiety. All but moderate-to-severe anxiety were more prevalent in the high (active) dose groups. Mild-to-moderate emotional distress was equally severe in low- and high-dose groups. Illusions, feeling abnormal, and feeling cold were also late AEs for one or two patients in the high-dose group. Severity was mostly mild to moderate. Examples of emotional distress and anxiety, as well as later resolution, were described by one patient as a "panic trip" with "pure fear of death," followed by relaxation and a "pleasant feeling" where "everything was light."

Discussion

This is the first overview of adverse events associated with the clinical use of MDMA and classic serotonergic psychedelics. Using a mixed-methods design, the review combined quantitative measures with qualitative descriptions of patient experiences. Based on this review, MDMA and psychedelics appear to be well tolerated in the treatment of various mental disorders. However, caution is necessary, as this conclusion is based on a limited number of relatively small studies with diverse study designs, varying quality, inconsistent AE assessment procedures, different types and doses of psychedelics, and patients with different disorders, often with prior experience using psychedelic drugs.

Acute AEs

The most common immediate physical AEs in the MDMA groups included fatigue, lack of appetite, feeling cold, thirst, jaw clenching, and sweating. Anxiety and difficulty concentrating were the most common immediate psychological AEs. The most common immediate AEs across psilocybin studies were moderate to severe anxiety, headache, and nausea. Immediate psychological AEs for psilocybin and LSD (e.g., paranoid thoughts, feeling trapped, illusions, feeling abnormal, psychological discomfort) mostly resolved during sessions but were sometimes severe. In one ayahuasca study, a participant temporarily experienced dissociation, fearing dying or losing control. Therapist interventions helped resolve this episode by calming the patient, highlighting the importance of trained, skilled, and trustworthy therapists. Therapist training is also important to familiarize therapists with the context of some AEs and effective ways to prevent or manage them.

Multiple qualitative studies described experiences that were occasionally terrifying, frightening, and confusing. In many cases, patients later viewed these experiences as therapeutically beneficial. Nausea and vomiting were common immediate AEs in ayahuasca studies. Except for a single case of irregular heartbeats in one MDMA-treated patient with PTSD, no drug-related AEs or SAEs required medical intervention.

Many of the physiological AEs described in this review are well known from research involving healthy individuals. Examples include nausea and vomiting with ayahuasca, headache after psilocybin use, and immediate, temporary elevated vital signs, jaw clenching, sweating, and lack of appetite with MDMA. Across studies and disorders, immediate psychological AEs, including anxiety, temporary paranoia, confusion, despair, and psychological distress, were commonly reported. These ranged from relatively mild to severe in some cases, and most resolved during the session.

Late AEs

The most common late AEs across MDMA studies were fatigue, headache, anxiety, difficulty concentrating, and low mood. A minority of patients reported anxiety, depression, panic, and suicidal thoughts in the days and weeks following treatment sessions. Low mood was a late AE present in both MDMA and placebo groups. Low mood is also commonly reported days after MDMA use in studies with healthy individuals and by recreational users. However, some studies do not report post-acute decreases in mood. Patients in control groups may have experienced low mood due to a lack of symptom relief and related disappointment; anxiety in the MDMA groups may be related to re-experiencing traumatic memories and painful emotions. Future research should more accurately characterize anxious states and low mood after MDMA treatment and assess whether these are a temporary part of the therapeutic process or a subacute negative effect.

One ayahuasca study reported hospitalizing four patients for a week, without further detailing how their "delicate condition" related to the effects of ayahuasca. Two studies reported unexpected negative reactions (post-traumatic flashback; severe worsening of anxiety and drug use relapse; worsening of depression) in three patients that occurred several weeks after psilocybin administration. These patients' conditions worsened and required further therapeutic intervention; one patient withdrew from the study. The authors noted that these reactions and the high incidence of anxiety, weeks after the session, may be related to the medical complexity and high number of co-occurring mental health conditions in this population. Despite the clinical severity of these complaints, they were not reported as SAEs, which risks underreporting potential negative consequences. However, the longer the time after treatment sessions, the less likely it is that events are still related to a specific treatment. Since individual responses and adverse reactions that may develop weeks or months after treatment cannot currently be predicted, it is crucial for researchers and clinicians to conduct longer-term follow-ups and report any unfavorable reactions.

Recently, cases were reported where study participants in an MDMA trial stated that they became suicidal after the trial concluded, including one case where a participant remained attached to a therapist and was subsequently abused. Such serious transgressions occur in the field and require strong action when they happen. The field of psychedelic treatments might be particularly at risk, as patients under the influence are vulnerable, suggestible, and highly dependent on therapists. Furthermore, therapists may have developed personal approaches to therapy that do not always adhere to professional conduct codes. Beyond this, it also highlights the importance of thoroughly and systematically monitoring adverse effects.

Suicidality

One study reported a suicide 11 days after a session with a placebo-like dose (1 mg) of psilocybin, stating it was unrelated to study participation. The results of a recently completed RCT comparing very low, medium, and high doses of psilocybin in patients with TRD were not included in this review. Overall AE rates seem roughly comparable to those summarized in this review. A preliminary publication reported suicidal thoughts, suicidal behavior, or intentional self-injury in 7 of 79 patients (8.8%) in the 25 mg group, 6 of 75 patients (8%) in the 10 mg group, and 4 of 79 patients (5%) in the 1 mg group; 14 instances in 9 patients were reported as SAEs. These numbers suggest higher suicidality rates in active dose groups (10 and 25 mg), but further interpretation awaits peer-reviewed publication. Suicidal thoughts and behavior were reported in three MDMA studies, although incidence was low, rates were lower or similar in MDMA compared to controls, and SAEs were only reported in the control groups. Nevertheless, careful assessment of treatment-related consequences is warranted. Being assigned to the inactive group, combined with despair and high expectations in patients with severe disorders, may contribute to feelings of hopelessness and subsequent suicidal thoughts or behavior. It is important to remember that mental disorders generally increase the risk of suicidal thoughts. Future studies need to carefully assess suicidal thoughts and explore ways to best support all patients long-term.

Interpreting adversity

Psychological or emotional "adverse events" can be ambiguous, subjective, context-dependent, and open to interpretation. The qualitative studies in this review illustrated how, in hindsight, patients viewed working through difficult feelings and emotions as therapeutically beneficial. This is supported by studies showing that users often transform challenging experiences into personally, morally, or spiritually meaningful ones. The shift in terminology—using "challenging experiences" instead of "bad trips"—indicates a growing acceptance that difficult experiences are an inherent part of psychedelic treatments and can be both distressing and therapeutically beneficial. More detailed accounts and the systematic use of specialized questionnaires can help interpret such experiences and their impact on treatment outcomes. This raises some important issues.

First, since the effects of psychedelics are thought to be significantly influenced by contextual factors, this may also apply to adverse effects. Treatment designs that reduce or minimize positive contextual components (e.g., time spent preparing patients, number of therapists, strength of the therapeutic relationship, time spent providing aftercare and integration) may increase the incidence of AEs. Second, some aspects of the definition of "AE" are not well suited to psychedelic treatment. By definition, AEs are undesired and something to be avoided. Categorizing challenging but potentially therapeutically beneficial experiences as AEs can therefore be counterproductive. For instance, framing "anxiety" as adversity may prompt patients to try to avoid difficult feelings and thoughts rather than engaging with them as part of a therapeutic process. Distinguishing adverse elements that should not be part of the therapeutic process (e.g., physical AEs such as headaches and nausea, or worsening of distressing symptoms such as suicidal thoughts and dissociation) from psychological experiences that are part of the therapeutic process is crucial for improving understanding of how psychedelic treatments work. To illustrate the complexity: vomiting, the most commonly reported AE in ayahuasca studies, is often considered a therapeutic element in both traditional and "neoshamanic" ayahuasca practices. In the case of anxiety, the severity and duration of the difficult episode may also predict suboptimal or even negative clinical outcomes. Careful screening, preparation, session monitoring, psychological support, and post-session integration are essential prerequisites to maintain a low incidence of AEs and maximize therapeutic effectiveness. Without these, young people and those with certain personality traits (neuroticism, emotional lability) or low in absorption, openness to experience, and acceptance, as well as those in an apprehensive, preoccupied, or confused mental state, may be more prone to challenging experiences that can worsen existing problems. Finally, recent articles have described how psychedelics may alter metaphysical or ontological beliefs. Such experiences can be highly meaningful but may also induce an "ontological shock," particularly when they conflict with previously held religious, personal, or spiritual beliefs. This introduces ethical and metaphysical challenges and a deep responsibility for therapists to mitigate potential adversity when patients confront profound metaphysical experiences.

Limitations

The current review has both strengths and limitations. An important strength is the systematic collection of studies and the combination of quantitative and qualitative studies. However, this review also has some important limitations. First, the included studies were mostly small and varied in design, drug dose, and demographics, making comparisons difficult and a meta-analysis impossible. Second, only a few studies assessed AEs consistently or systematically. Unfortunately, this is a general problem in psychopharmacology, not unique to psychedelic drug studies. Most studies relied on spontaneously reported and/or therapist-observed AEs. Spontaneous reporting depends on individual and contextual factors that allow patients to recall or discuss negative experiences. Therapist observation is similarly ambiguous, as observations may be influenced by their subjective perception and interpretation. Five pilot studies did not report any AEs. While it is possible, though unlikely, that no AE occurred, not reporting any AE is similar to important omissions in early psychedelic research, where researchers often did not fully disclose the frequency and severity of negative reactions. Third, there are not enough placebo-controlled trials to draw strong conclusions about the causal relationship between psychedelic treatment and AEs; effectively blinding study participants to their treatment condition is particularly difficult in psychedelic drug studies, especially given the high percentages of patients who have had previous experience with a psychedelic. Also, due to small sample sizes, studies may have missed less frequent AEs, hindering broader application as some AEs may only become apparent in larger populations. All studies in this review refer to a total of 598 unique patients, of whom 521 received an active dose. Fourth, the severity and onset timing of AEs (particularly for late AEs) were reported in only a few studies, making it difficult to adequately interpret reported AEs. Next, most trials had strict participation criteria, with extensive screening and exclusion of patients with almost any co-occurring mental disorder. This includes a personal or family history of psychotic, bipolar, dissociative identity, eating, and substance use disorders, and active suicidal thoughts, all of which are frequently found in patients with more severe mental disorders. Sixth, many studies included patients with previous experience with the experimental drug; up to 56% for MDMA and roughly half of all psilocybin study participants had prior experience. The intensity of drug effects and experienced difficulties appear to be lower and less severe in those with past drug experience, limiting the generalizability to broader patient populations. Additionally, minority groups are underrepresented in clinical research with psychedelics. Therefore, it is possible that patients who are new to psychedelics, have a different socio-cultural background, and more severe forms of mental illness may have more intense experiences or more difficulty coping with potentially disruptive events. However, the only study in this review that compared lifetime psychedelic use and the degree of experienced challenge within the session found no significant correlations, suggesting that previous experience does not necessarily predict the occurrence of AEs in clinical populations or contexts. Seventh, while qualitative studies helped provide context for some of the immediate challenging experiences, these studies were not always specifically designed to explore such effects. Furthermore, qualitative studies are unevenly distributed and relatively scarce; none were on ayahuasca, only one on MDMA and LSD, and eight on clinical psilocybin studies. Finally, any conclusion on the frequency and interpretation of AEs should consider the patient's mindset and the treatment setting. More information about these factors should be provided when reporting and interpreting the results of clinical trials with psychedelics.

Conclusions

This first review of immediate and late adverse events in treatment with serotonergic psychedelics and MDMA, combining quantitative and qualitative studies, shows that all compounds immediately induce temporary headaches, nausea, and anxiety. Anxiety covers a wide range of psychologically challenging events that may also benefit the therapeutic process. Distinguishing truly adverse events from potentially beneficial effects is complex but crucial for improving understanding of psychedelic treatments. Serious adverse events appear largely absent based on the included studies, as were lasting physiological side effects. Suicidal thoughts and behavior were rare but did occur; these represent a serious psychiatric emergency, emphasizing the importance of investigating which patients are most at risk and how best to reduce the likelihood of their occurrence. Overall, adverse events were inconsistently defined and inadequately assessed. Future studies should describe the timing and severity of effects more extensively, for example, using specialized scales. Many studies included patients with prior psychedelic experience, which may bias results and limit generalizability. Qualitative research can also add nuance by detailing and understanding the meaning of challenging experiences. Full transparency about adverse events is a responsibility of clinicians, particularly in a developing field driven by the enthusiasm of pioneering researchers. Understanding the full range of unpleasant, potentially harmful, and transformative treatment-related events is crucial to inform future therapists who may otherwise be inadequately prepared to handle challenging and potentially destabilizing patient experiences, especially in larger groups of patients with more complex and potentially co-occurring conditions.

Introduction

Clinical research using psychedelic substances has shown promising results for treating various mental health conditions. However, many of these studies involved a small number of carefully selected participants. With the completion of initial large-scale clinical trials and ongoing studies exploring new applications, some of these compounds appear close to becoming approved treatments.

Using psychedelics, especially in uncontrolled settings, can lead to immediate negative effects. These might include anxiety, panic, extreme sadness, paranoia, or dangerous behaviors. Such experiences could contribute to long-lasting psychological problems. While most negative outcomes have been reported from non-medical use, and small clinical trials have shown that these substances are generally safe and well-tolerated, administering them to larger, more diverse patient populations with other health issues might uncover unexpected negative effects. Additionally, severe reactions to a drug can lead patients to stop treatment. Unresolved or difficult experiences might also result in ongoing negative outcomes. Therefore, it is important to identify the full range of adverse reactions to psychedelic drugs, particularly in vulnerable patients who have not responded to other treatments. The scientific literature on this topic has not been thoroughly reviewed since 1984. Earlier reviews, like one from 1960 that surveyed clinicians who had treated about 5,000 patients (mostly with LSD), suggested that negative effects were rare, except in some patients with schizophrenia, and that the drugs were generally safe when administered carefully. A 1984 review noted a range of adverse reactions, from acute, short-term panic during administration to temporary psychoses lasting several days, and even recurring flashbacks or chronic psychotic cases. Flashbacks are now a recognized diagnostic condition called "hallucinogen persisting perception disorder" (HPPD), which is listed in the DSM-V, though it is reported mostly in cases of frequent non-medical use of certain psychedelics and may be linked to existing psychiatric conditions. However, flashbacks are often described as mild or even pleasant. A recent review explored how much of the publicly discussed harms are based on scientific evidence versus anecdotes.

Evaluating adverse events (AEs) is difficult for several reasons. AEs are not always planned for in studies, the range of possible reactions is wide, reporting can be inconsistent, and terms like "side effects," "toxic effects," and "adverse drug reactions" are often used interchangeably. This complexity is further increased by the highly varied and context-dependent subjective effects that psychedelics can produce.

As more patients with mental health conditions may soon be treated with psychedelics, a complete understanding of both their benefits and their potential negative effects is needed. This information helps determine the overall benefits and risks, and decide which patients are most likely to benefit or experience harm. This paper systematically reviews any adverse events that occur during or after psychedelic treatments involving classic hallucinogens (such as psilocybin, LSD, ayahuasca) and entactogens (such as MDMA) in patients. Even with their different pharmacological properties, MDMA and serotonergic hallucinogens are often grouped together as "psychedelics" and used in similar therapeutic settings. Adverse events related to ketamine and ibogaine have been reviewed in other publications.

Methods

This review takes an exploratory approach rather than trying to confirm a specific hypothesis, due to the newness of this field. It was not pre-registered. The systematic review and data synthesis follows a detailed design, using established guidelines for extracting, organizing, and combining findings from both quantitative and qualitative studies. In this review, an adverse event (AE) refers to any unfavorable or harmful event that first occurs during or after the administration of a classic psychedelic or MDMA, regardless of whether it is directly related to the treatment. Acute AEs are defined as those occurring during or on the same day as a psychedelic or MDMA session. Late AEs are those that emerge after the day of the session. While other terms like "treatment-emergent AEs" (TEAEs) exist, this review focuses only on the distinction between acute and late AEs.

Selection criteria

The review included all quantitative and qualitative clinical studies published since 2000 that described treating patients with a mental disorder using classic serotonergic psychedelics (like psilocybin, LSD, mescaline, or ayahuasca) or entactogenic drugs (like MDMA or MDA). This included open-label studies, case reports or series, and randomized controlled trials (RCTs). The review excluded systematic reviews, surveys, secondary analyses, and studies involving healthy volunteers. The goal was to assess all possible adverse reactions in patients with a diagnosed mental disorder who received treatment with these compounds. Therefore, no specific exclusion criteria were set beforehand to allow for the broadest possible inclusion and avoid missing any potential unfavorable outcomes.

Search methods

A systematic search of the PubMed/Medline, EMBASE, and PSYCinfo databases was conducted on July 28, 2021. The search combined specific terms for psychedelic compounds (e.g., psilocybin, lysergic acid diethylamide, MDMA) with terms for study types (e.g., clinical trial, qualitative), while excluding animal studies. ClinicalTrials.gov was also searched to find corresponding publications. These systematic searches were supplemented by manually checking references in relevant articles.

Quality assessment

A formal, in-depth quality assessment was not considered essential because even high-quality studies with low risk of bias may not report adverse events thoroughly. Instead, the review specifically assessed quality criteria relevant to its aim: the methods used to monitor or report AEs, the study's inclusion and exclusion criteria, and the percentage of participants with prior experience using the drug. The Critical Appraisal Skills Programme (CASP) checklist was used to assess the methodological rigor of qualitative studies.

Data extraction and synthesis

All descriptive and quantitative information about acute and late AEs, the timing of their assessment, and serious AEs (SAEs) was extracted from the main text and supplementary files of the publications. The prevalence of AEs was recalculated as a percentage of total participants. Qualitative articles were examined for themes and descriptions related to AEs, which were then used to support, illustrate, and provide context for the findings from quantitative studies.

Results

The initial search yielded 5,640 articles. After removing duplicates, 3,709 titles and abstracts were screened. Following this screening, the full texts of 61 articles were assessed for eligibility, and 44 articles were ultimately included. The review also found 596 registered trials, and after selection, 24 completed trials were included.

The included studies were published between 2006 and 2021 and involved a total of 598 unique participants, with 521 receiving an active dose. Study sizes ranged from 1 to 105 participants, and participant ages ranged from 25 to 59 years. Multiple publications referring to the same study were combined. All qualitative studies described sub-samples from quantitative studies included in this review. There was significant variation in study design, substances, dosages, and mental disorders studied. Four quantitative and two qualitative studies did not report on adverse events. The review will now describe the main results for each compound, organized by disorder. Tables summarizing the included studies and reported AEs were provided in the original document.

MDMA

Sixteen studies were selected for MDMA, including one qualitative study, involving a total of 266 patients. Ten of these publications described MDMA-assisted treatment for PTSD (214 patients), including seven RCTs, one open-label trial, one case series, and one qualitative study. Other RCTs reported on treating end-of-life anxiety, social anxiety in adults with autism, and tinnitus. One open-label study focused on alcohol use disorder. Active doses ranged from 50 to 125 mg, sometimes with an optional half-dose supplement.

Regarding study quality, 11 MDMA studies reported AEs only when spontaneously mentioned by participants, and one study did not report AEs at all. Two RCTs systematically assessed AEs, and one study reported AE severity using a specific scale. Six studies did not report whether participants had used MDMA or "ecstasy" before the study. In the other nine quantitative studies, prior MDMA exposure varied from none to 56%. The qualitative study was of medium to high quality but did not report on AEs.

Physiological measurements in nine studies consistently showed mild, temporary, and statistically significant increases in blood pressure, body temperature, and heart rate during MDMA sessions. None of these changes required medical intervention. One MDMA-related serious adverse event (SAE) was reported: a participant experienced an increase in premature ventricular contractions (which were also present at baseline), requiring one night of hospital monitoring. This issue resolved spontaneously.

For patients with PTSD, the most common immediate physical AEs were jaw clenching, muscle tightness, headaches, nausea, fatigue, and lack of appetite. Anxiety was the most common immediate psychological AE. Most immediate AEs occurred more often in the MDMA groups than in placebo groups, though there were some exceptions. Immediate AEs were generally mild to moderate in severity. There was no clear relationship between MDMA dose and AE occurrence in studies that tested different doses. Two studies reported SAEs. In one, two patients in the control group reported suicidal behavior or thoughts requiring hospitalization. In another, three out of four SAEs were deemed unrelated to the treatment; one patient developed an exacerbation of a pre-existing heart condition, which resolved without evidence of cardiac disease and was considered potentially related to the drug. Common late AEs included fatigue, lack of appetite, low mood, insomnia, increased need for sleep, irritability, headache, difficulty concentrating, and anxiety. A minority of patients reported complaints up to two months after the final session, with anxiety, low mood, and panic attacks sometimes being more frequent in the MDMA group.

For end-of-life anxiety, an RCT reported jaw clenching, thirst, dry mouth, perspiration, and headache as immediate AEs. Fatigue, increased need for sleep, insomnia, anxiety, jaw clenching, and low mood were the most common late AEs. All except low mood occurred more frequently in the MDMA group. Anxiety, depressed mood, and dissociation were reported only in the MDMA group.

In social anxiety studies, immediate AEs like anxiety, difficulty concentrating, fatigue, headache, and sensitivity to cold were more common in the MDMA group. Common late AEs included fatigue, headache, difficulty concentrating, low mood, and increased need for sleep; all except the latter two were more common in the MDMA group. Two of eight patients in the MDMA group and one of four in the placebo group reported depressed mood, suicidal thoughts, and panic attacks after the first drug session. Studies on tinnitus and alcohol use disorder did not report on adverse events.

Psilocybin

Twenty articles were included for psilocybin, describing treatment in 257 patients. These included six RCTs, five articles describing four open-label studies, one dose-escalation study, and eight qualitative studies. Studies investigated psilocybin for end-of-life anxiety, major depressive disorder (MDD), migraine headaches, obsessive-compulsive disorder, treatment-resistant depression (TRD), alcohol use disorder, tobacco smoking, and demoralization in older AIDS-survivors. Active dosages varied from 10 to 30 mg or 0.1 to 0.4 mg/kg of psilocybin per session.

Two studies did not report on AEs. Most others relied on spontaneous patient reports and therapist observations, sometimes supplemented by interviews or questionnaires. Three studies reported AE severity. In studies assessing dose-effect relationships, one found some psychological AEs more prevalent at a moderate dose (20 mg) than a high dose (30 mg), and another reported more post-treatment headaches in the higher dose group. One study reported a much higher number of distinct AEs than others. Most quantitative studies (with one exception) reported on prior psychedelic use, ranging from 20% to 100% of participants, with an average of 50%. The quality of the eight qualitative studies was high, though one did not report AEs.

Elevated blood pressure was the most common physiological effect in the psilocybin group, significant in three studies. Increased heart rate was significant in one study but not others. None of these vital sign elevations required medical intervention. Self-limiting severe hypertension occurred in 4 of 18 patients in one study, linked to severe anxiety and resolved by therapist reassurance. Another study noted one patient's blood pressure exceeding protocol criteria, which resolved without intervention.

For depressive disorders, an open-label study for TRD reported anxiety, confusion, and nausea as the most common immediate AEs. Anxiety mostly occurred before and at the start of sessions, while confusion occurred during the peak. Severity was mild to moderate, except for one case of severe anxiety during a high-dose session. Both MDD RCTs reported headaches. One RCT noted moderate to strong emotional and psychological AEs, including crying, sadness, emotional/physical suffering, grief, and isolation. One participant experienced confusing traumatic memories, which worsened depression for several weeks. Another became incommunicative during a session. A patient described a fearful response to sensory input, afraid to open their eyes, while another described deep sadness, grief, and a process of unblocking. Post-treatment headache was the only reported late AE. In one study, 67% of patients rated a psilocybin session as one of their top five most psychologically challenging experiences.

For end-of-life anxiety, two studies reported immediate and temporary AEs in psilocybin groups, including physical discomfort (headache, nausea) and psychological discomfort (anxiety, transient thought disorders, suicidal thoughts). One study did not report AEs. Headache was the only reported late AE. One patient died by suicide 11 days after a low psilocybin dose, though authors concluded it was unrelated to the study. Qualitative accounts described terrifying and disorienting experiences that resolved with therapist support.

In substance use disorder studies, one reported strong to extreme anxiety and feeling trapped as immediate AEs, and headache as a late AE. Another reported few AEs, though qualitative studies described frightening experiences or chaotic states that eventually led to peacefulness.

For migraines, the most reported immediate AEs were nausea, anxiety, and lightheadedness; headache and migraine attacks were the only late AEs. All were more prevalent in the psilocybin group. In the demoralization study, the most common immediate AEs were anxiety, nausea, and headache. Most of these were temporary, and nearly half were moderate to severe. Two unexpected late AEs occurred: one patient had a post-traumatic flashback, preventing work for two days, and another experienced severe anxiety exacerbation followed by methamphetamine use relapse, leading to study withdrawal.

Ayahuasca

Three studies were included for ayahuasca, totaling 48 patients: one RCT for social anxiety disorder, one RCT for TRD patients, and one open-label study for MDD patients. All involved a single administration.

None of the studies systematically assessed AEs, and only one reported acute AEs spontaneously or by therapist observation. All participants were new to ayahuasca, and most to other serotonergic psychedelics.

No significant increases in heart rate or blood pressure were observed in the studies that reported physiological parameters. Nausea, vomiting, diarrhea, and anxiety were more common immediate AEs in the ayahuasca group. In the TRD study, four patients remained hospitalized for seven days due to a "more delicate condition," without further details. In the social anxiety study, one participant experienced an intense episode of temporary fear of dying or going crazy, distress, and dissociation, which resolved with therapist reassurance. No AEs were reported in the control group of this study. The open-label study only reported vomiting as an immediate AE. No studies reported late AEs.

LSD

Four studies involving 27 patients were included for LSD: one RCT with a qualitative component describing end-of-life anxiety treatment, one observational study on compassionate use of LSD and/or MDMA in group therapy for PTSD and/or MDD, and one case report on LSD treatment for a complex personality disorder. Active doses of 100–200 μg were used.

AEs were collected through spontaneous reporting by patients and therapist observations. In the compassionate group therapy study, 39% of participants had prior drug experience. Only one participant (8%) in the RCT had prior experience. The qualitative study was of medium to high quality. AE severity was reported only in the RCT.

No significant differences in heart rate or blood pressure were observed between groups. Immediate, temporary AEs included illusions, feeling cold, feeling abnormal, and anxiety. All but moderate-to-severe anxiety were more common in the high-dose groups. Mild-to-moderate emotional distress was equally severe in low- and high-dose groups. Illusions, feeling abnormal, and feeling cold were also reported as late AEs for a few patients in the high-dose group. Severity was mostly mild to moderate. Patient accounts described initial panic and fear of death that eventually led to relaxation and pleasant feelings.

Discussion

This review, the first to examine adverse events with clinical MDMA and classic serotonergic psychedelics by combining quantitative and qualitative studies, suggests that these compounds are generally well tolerated for various mental disorders. However, caution is warranted. This conclusion is based on a limited number of relatively small studies with varying designs, inconsistent AE assessment, different psychedelic types and doses, and patients with high rates of prior drug experience.

The most common immediate physical AEs in MDMA groups were fatigue, lack of appetite, feeling cold, thirst, jaw clenching, and perspiration. Anxiety and difficulty concentrating were the most common immediate psychological AEs. For psilocybin studies, the most common immediate AEs were moderate to severe anxiety, headache, and nausea. Immediate psychological AEs for psilocybin and LSD (e.g., paranoid thoughts, feeling trapped, illusions, feeling abnormal, psychological discomfort) typically resolved during sessions but were sometimes severe. In one ayahuasca study, a participant temporarily felt they were dying or going crazy and dissociated. Therapist interventions, emphasizing their importance, helped resolve this episode. Many qualitative studies described terrifying or confusing experiences that patients later viewed as therapeutically beneficial. Nausea and vomiting were common immediate AEs in ayahuasca studies. With the exception of one case of premature ventricular contractions in an MDMA-treated PTSD patient, no drug-related AEs or serious AEs required medical intervention. Many physiological AEs described are well-known from studies in healthy individuals. Across studies and disorders, immediate psychological AEs like anxiety, temporary paranoia, confusion, despair, and psychological distress were frequently reported, ranging from mild to severe, with most resolving during the session.

The most common late AEs across MDMA studies were fatigue, headache, anxiety, difficulty concentrating, and low mood. A minority of patients reported anxiety, depression, panic, and suicidal thoughts in the days and weeks following treatment. Low mood was observed in both MDMA and placebo groups, a known phenomenon after MDMA use in healthy individuals. However, some studies did not report post-acute mood decreases. Patients in control groups might experience low mood due to lack of symptom relief, while anxiety in MDMA groups could relate to re-experiencing traumatic memories. Future research should better characterize anxious states and low mood after MDMA treatment to determine if they are a temporary part of the therapeutic process or a subacute adverse effect. One ayahuasca study reported hospitalizing four patients for a week due to a "more delicate condition," without further explanation. Two psilocybin studies reported unexpected late adverse reactions in three patients (post-traumatic flashback, severe anxiety exacerbation and drug use relapse, depression worsening) several weeks after administration, requiring further therapeutic intervention and leading to one patient withdrawing. Authors noted these might be related to the medical complexity and other psychiatric conditions in this population. Despite their severity, these complaints were not reported as serious AEs, which risks underreporting potential negative consequences. However, events occurring long after treatment are less likely to be directly related to it. Since individual responses and adverse reactions weeks or months after treatment cannot currently be predicted, researchers and clinicians must conduct longer-term follow-ups and report all unfavorable reactions. Recent reports of serious ethical violations, including a participant becoming suicidal and abused after an MDMA trial, highlight the extreme vulnerability of patients under the influence and the critical need for rigorous adverse effect monitoring and adherence to professional codes of conduct.

One study reported a suicide 11 days after a placebo-like dose of psilocybin, concluding it was unrelated to the study. Preliminary, non-peer-reviewed results from a large psilocybin RCT in TRD patients reported suicidal thoughts, behavior, or self-injury in a higher percentage of patients in the active dose groups (10 mg and 25 mg) compared to the 1 mg group, with some instances reported as serious AEs. Suicidal ideation and behavior were rare in MDMA studies, with rates similar to or lower than control groups, and serious AEs only reported in control groups. Nevertheless, thorough assessment of treatment consequences is essential. Being assigned to an inactive group, combined with high expectations in patients with severe disorders, may contribute to hopelessness and suicidal thoughts. It is important to remember that mental disorders generally increase the risk of suicidal thoughts. Future studies must carefully assess suicidal ideation and explore ways to best support all patients long-term.

Psychological or emotional "adverse events" can be ambiguous, subjective, context-dependent, and open to interpretation. Qualitative studies in this review showed that patients often retrospectively considered working through difficult feelings and emotions therapeutically beneficial. This is supported by studies showing that users often transform challenging experiences into personally meaningful ones. The shift in terminology from "bad trips" to "challenging experiences" indicates a growing acceptance that difficult experiences are inherent to psychedelic treatments and can be both distressing and beneficial. This raises important issues. First, since psychedelic effects are heavily influenced by context, this may also apply to adverse effects. Treatment designs that minimize positive contextual components (e.g., patient preparation, therapist support, aftercare) may increase the incidence of AEs. Second, parts of the definition of "AE" are ill-suited for psychedelic treatment. AEs are by definition undesired. Categorizing challenging but potentially therapeutic experiences as AEs can be counterproductive. For instance, labeling "anxiety" as an AE might make patients avoid difficult feelings rather than engaging with them therapeutically. Distinguishing between genuine adverse elements (e.g., physical AEs like headaches, or worsening distressing symptoms like suicidal thoughts or dissociation) and psychological experiences that are part of the therapeutic process is crucial for understanding how psychedelic treatments work. For example, vomiting, common in ayahuasca studies, is often seen as a therapeutic element in traditional practices. For anxiety, its severity and duration may predict suboptimal or negative outcomes. Careful screening, preparation, session monitoring, psychological support, and post-session integration are essential to maintain a low incidence of AEs and maximize therapeutic efficacy. Without these, individuals with certain personality traits or mental states may be more prone to challenging experiences that could worsen existing problems. Finally, recent articles describe how psychedelics can alter metaphysical or spiritual beliefs. Such experiences, though meaningful, can also induce an "ontological shock," especially if they conflict with prior beliefs. This introduces ethical and metaphysical challenges, placing a deep responsibility on therapists to manage potential adverse outcomes when patients face profound metaphysical experiences.

This review has both strengths and limitations. A key strength is its systematic collection of studies and combination of quantitative and qualitative findings. However, a major limitation is that included studies were mostly small, varied in design, drug type and dose, and patient demographics, making comparisons difficult and preventing meta-analysis. Few studies assessed AEs consistently or systematically, often relying on spontaneous reports or therapist observations, which can be ambiguous. Five pilot studies reported no AEs, which is unlikely and echoes past omissions in early psychedelic research regarding the full disclosure of negative reactions. There are not enough placebo-controlled trials to draw strong conclusions about the causal link between psychedelic treatment and AEs, partly because effectively blinding participants in psychedelic studies is hard, especially given the high percentage of patients with prior psychedelic experience. Due to small sample sizes, less frequent AEs might have been missed, limiting generalizability to larger populations. Severity and timing of AEs were rarely reported, hindering interpretation. Most trials had strict participation criteria, excluding patients with many comorbid mental disorders or active suicidal thoughts. Many studies included patients with prior drug experience (up to 56% for MDMA, about half for psilocybin), which might bias results since experienced users often report less severe difficulties, limiting generalizability to broader patient populations. Additionally, minorities are underrepresented. It is possible that psychedelic-naïve patients from different socio-cultural backgrounds or with more severe mental illness might have more intense or difficult experiences. However, one study found no correlation between prior psychedelic use and the degree of challenge experienced in clinical settings. Lastly, while qualitative studies helped contextualize acute challenging experiences, they were not always designed to specifically explore such effects and are unevenly distributed across drugs. Any conclusions on AE frequency and interpretation must consider the patient's mindset ("set") and the treatment environment ("setting"). More information about "set and setting" should be provided when reporting clinical trial results with psychedelics.

Conclusions

This initial review of immediate and late adverse events in treatments with serotonergic psychedelics and MDMA, combining quantitative and qualitative studies, indicates that all compounds acutely induce temporary headaches, nausea, and anxiety. Anxiety can encompass a wide range of psychologically challenging experiences that may also be beneficial to the therapeutic process. Separating true adverse events from potentially beneficial effects is complex but vital for improving understanding of psychedelic treatments. Serious adverse events appear largely absent based on the included studies, as were lasting physiological side effects. Suicidal ideation and behavior were rare but did occur, representing a serious psychiatric emergency. This highlights the importance of investigating which patients are most at risk and how best to reduce the likelihood of these events. Overall, adverse events were inconsistently defined and inadequately assessed across studies. Future studies should describe the timing and severity of effects more thoroughly, for instance, by using specialized questionnaires. Many studies included patients with prior psychedelic experience, which may bias results and limit their applicability to a broader population. Qualitative research can also add nuance by detailing and explaining the meaning of challenging experiences. Full transparency about adverse events is a responsibility for clinicians, especially in a new field driven by pioneering researchers. Understanding the complete range of unpleasant, potentially harmful, and transformative treatment-related events is crucial to prepare future therapists, who might otherwise be inadequately equipped to handle challenging and potentially destabilizing patient experiences, particularly in larger groups of patients with more complex and potentially co-occurring conditions.

Introduction

Recent clinical research involving psychedelic substances has shown promising positive results for various mental health conditions. However, many of these studies have been small, involving a limited and specific group of participants. With some larger, randomized clinical trials nearing completion and new uses being explored, certain psychedelic compounds may soon be approved for medical use.

The use of psychedelics, especially outside of medical settings, can lead to immediate negative effects. These include anxiety, panic, extreme sadness, paranoia, and risky behaviors. Such experiences might also contribute to long-term psychological issues. While these problems are mostly linked to non-medical use, and small clinical trials have shown the drugs can be safe and well-tolerated, giving them to larger, more diverse groups of patients with multiple health issues could reveal unexpected negative effects. Additionally, unwanted drug reactions can cause patients to stop treatment, and difficult experiences that are not properly dealt with might lead to lasting problems. Therefore, it is important to identify all possible negative reactions to psychedelic drugs, especially in vulnerable patients who have mental disorders that are hard to treat.

There has not been a thorough review of this topic since 1984. An earlier review from 1960, based on clinician surveys from about 5,000 patients (mostly treated with LSD), concluded that negative effects were rare, except in some patients with schizophrenia, and that the drugs were generally safe when used carefully. A 1984 review found that negative reactions varied, from temporary panic during use, to short-term psychosis lasting several days, and even recurring flashbacks or long-term psychotic conditions that resist treatment. Flashbacks were later included in a diagnostic category called "hallucinogen persisting perception disorder" (HPPD). HPPD is sometimes reported, usually in people who frequently use certain psychedelics or MDMA outside of medical settings, and it may be linked to existing mental health issues. However, flashbacks are often described as mild and pleasant. A recent review looked at common negative effects discussed publicly to see which ones had scientific support.

Assessing negative effects is difficult for many reasons. These effects are not always clearly defined beforehand, they can vary widely, reporting can be inconsistent, and the terms used are often confusing. This challenge is made greater by the highly variable and personal effects that psychedelics can have, which depend heavily on the context of their use.

As more patients with mental disorders may soon be treated with psychedelics, a complete overview of both their benefits and negative effects is needed. This will help doctors assess risks and benefits, and decide which patients are most likely to benefit or not. This paper aims to systematically review any negative effects that occur during or after psychedelic treatments with classic hallucinogens (such as psilocybin, LSD, ayahuasca) and MDMA in patients. While MDMA and classic hallucinogens have different pharmacological properties, they are often grouped together as "psychedelics" and used under similar treatment conditions. Negative effects of other psychedelics like ketamine and ibogaine have been reviewed elsewhere.

Methods

Because this field of study is relatively new, this review uses an exploratory approach rather than trying to confirm specific hypotheses. This systematic review and data analysis follows a step-by-step design, using established guidelines to gather, organize, and summarize findings from both quantitative (number-based) and qualitative (experience-based) studies. The term "adverse event" (AE) is used for any unfavorable or harmful event that first occurs during or after the administration of a classic psychedelic or MDMA, regardless of whether it is directly related to the treatment. In this review, acute AEs are those that happen during or on the day of a psychedelic or MDMA session, while late AEs are those that appear after the session day.

All quantitative and qualitative clinical studies published since 2000 were included. These studies describe the treatment of patients with a mental disorder using classic serotonergic psychedelics (like psilocybin, LSD, mescaline, ayahuasca) or entactogenic drugs (like MDMA, MDA). This includes open-label studies, case reports or series, and randomized controlled trials (RCTs). Reviews, surveys, secondary analyses, and studies with healthy volunteers were excluded. The goal was to assess possible negative reactions in patients diagnosed with a mental disorder who received these compounds. No specific exclusion criteria were set beforehand to ensure the broadest possible inclusion and to avoid missing any potential unfavorable outcomes.

A systematic search was performed on July 28, 2021, using PubMed/Medline, EMBASE, and PSYCinfo databases. The search combined terms for psychedelic compounds (e.g., psilocybin, LSD, MDMA) and study types (e.g., clinical trial, qualitative), while excluding animal studies. ClinicalTrials.gov was also searched to find related publications, and additional studies were found by hand-searching and checking reference lists.

A thorough formal assessment of study quality was not considered essential, as even high-quality studies with low bias might not report adverse events well. Instead, specific quality criteria important for this review's aim were assessed: the methods used to monitor or report AEs, study inclusion and exclusion criteria, and the percentage of participants with previous experience using the drug. A checklist was used to assess the rigor of qualitative studies.

All descriptive and quantitative information about acute and late AEs, when they were assessed, and serious AEs (SAEs) was extracted from the main text and supplementary files of the publications. AE prevalence was recalculated as a percentage of total participants. Qualitative articles were carefully examined for themes and descriptions related to AEs, which were then used to support, illustrate, and provide context for findings from quantitative studies.

Results Overview

The initial search yielded 5,640 articles. After removing duplicates, 3,709 titles and abstracts were reviewed. Disagreements about inclusion were resolved through discussion among multiple authors. After this screening, 61 articles were fully reviewed for eligibility, and 44 were included. Additionally, out of 596 registered trials, 24 completed trials were included.

The included studies were published between 2006 and 2021 and involved a total of 598 unique participants, with 521 receiving an active dose. Study sizes ranged from 1 to 105 participants (average 23), with average ages from 25 to 59 years. Publications referring to the same study were combined. All qualitative studies described sub-groups from quantitative studies included in this review. There was significant variation in study design, substances used, dosages, and the types of disorders treated. Four quantitative and two qualitative studies did not report on adverse events.

MDMA Research Findings

Sixteen studies were included, with one being a qualitative study, involving a total of 266 treated patients. Ten of these studies focused on MDMA-assisted treatment for PTSD, including seven randomized controlled trials (RCTs), one open-label trial, one case series, and one qualitative study. Other RCTs looked at MDMA for end-of-life anxiety, social anxiety in adults with autism, and tinnitus. One open-label study focused on alcohol use disorder. Active doses typically ranged from 50 to 125 mg, sometimes followed by an optional half-dose.

For quality assessment, 11 of the MDMA studies only reported adverse events if participants mentioned them spontaneously. One study did not report any adverse events. Two RCTs systematically assessed adverse events, and one study used a specific scale to report their severity. Six studies did not mention if participants had previously used MDMA or "ecstasy." In the other nine quantitative studies, prior MDMA exposure varied from none to 56%. The qualitative study was of medium to high quality but did not report on adverse events.

Nine studies measured blood pressure, body temperature, and heart rate during MDMA sessions. All reported mild, temporary, and statistically significant increases in these measures. None required medical intervention. One serious adverse event (SAE) potentially related to MDMA was reported: a participant experienced an increase in irregular heartbeats, which were also present before the study. This required one night of hospital monitoring and resolved on its own.

MDMA Adverse Reactions

For PTSD: The most common immediate physical adverse events were jaw clenching or tight muscles, headaches, nausea, fatigue, and lack of appetite. Other immediate effects included feeling cold, thirst, dizziness, sweating, restlessness, and body pains. Anxiety was the most common immediate psychological adverse event. Most immediate adverse events happened more often in the MDMA groups than in the placebo groups, though in one RCT, anxiety and fatigue were slightly more common in the placebo group. Immediate adverse events were mild to moderate in severity in the single study that reported it. There was no clear link between the MDMA dose and the occurrence of adverse events in two PTSD studies that tested different doses. Serious adverse events (SAEs) were reported in two studies: in one, two patients reported suicidal thoughts or behavior leading to hospitalization, but these occurred only in the control group. In another study, three out of four SAEs were considered unrelated to the treatment. One patient developed worse pre-existing irregular heartbeats, which resolved without heart disease, and this was considered possibly related to treatment.

One case study described a participant experiencing strong emotional reactions, like crying and grief, during sessions, along with physical reactions such as muscle tightening and sweating, especially when reviewing traumatic memories. The participant later reflected, "There’s no easy fix—I need to work through the darkness."

Common late adverse events included fatigue, lack of appetite, low mood, insomnia, needing more sleep, increased irritability, headache, difficulty concentrating, and anxiety. In one study, rates of fatigue, insomnia, and headaches were somewhat higher in the control group. A small number of patients in this group reported some complaints up to two months after the final session, including anxiety (17 of 72 patients), depressed or low mood (6 of 72 patients), and panic attacks (4 of 72 patients), with higher rates in the MDMA group than the control group.

For End-of-Life Anxiety: A study on end-of-life anxiety reported jaw clenching, thirst, dry mouth, sweating, and headache as immediate adverse events. Fatigue, needing more sleep, insomnia, anxiety, jaw clenching, and low mood were the most common late adverse events. Except for low mood, all these effects occurred more often in the MDMA group. Anxiety, depressed mood, and feeling disconnected were reported only in the MDMA group; insomnia was reported in both groups.

For Social Anxiety: A study on social anxiety in adults with autism found higher rates of immediate adverse events in the MDMA group compared to the placebo group. These included anxiety, difficulty concentrating, fatigue, headache, and sensitivity to cold. The most common late adverse events were fatigue, headache, difficulty concentrating, low mood, and needing more sleep. All late adverse events except low mood and needing more sleep were more common in the MDMA group. Two of eight patients in the MDMA group and one of four in the placebo group reported depressed mood, suicidal thoughts, and panic attacks after the first drug session.

For Tinnitus and Alcohol Use Disorder (AUD): The studies on tinnitus and AUD treatment did not report on adverse events.

Psilocybin Research Findings

Twenty articles were included, describing psilocybin treatment in 257 patients. This included six randomized controlled trials (crossover studies), five articles detailing four open-label studies, one single-group dose-escalation study, and eight qualitative studies. RCTs focused on end-of-life anxiety and major depressive disorder (MDD), and migraine headaches. The dose-escalation study looked at obsessive-compulsive disorder. Open-label studies investigated treatment-resistant depression (TRD), alcohol use disorder (AUD), tobacco smoking, and low morale in older AIDS survivors. Active dosages varied from 10 to 30 mg or 0.1 to 0.4 mg/kg of psilocybin per session. Only two of six studies that used different psilocybin doses reported on adverse effects per dosage. One study described the timing of adverse events per session. Qualitative studies detailed treatment experiences among patients with end-of-life anxiety, TRD, and substance use disorder.

Two studies did not report on adverse events. Most other studies relied on patients' spontaneous reports and therapists' observations. Some also used post-treatment interviews, headache ratings, and structured questionnaires. Three studies reported on the severity of adverse events. In studies that assessed dose-effect relationships, one found a higher rate of some psychological adverse events at a moderate dose (20 mg) compared to a high dose (30 mg), and another reported more post-treatment headaches in the higher dose group. One study reported a much higher number of distinct adverse events than all other studies.

With one exception, all quantitative studies reported on prior psychedelic use, which ranged from 20% to 100% of participants, averaging 50% of patients having used psilocybin or other psychedelics before the study. One study reported an average of 0.8 lifetime psychedelic experiences per group, while in another, 39% of patients had used psilocybin 10 or more times, with 22% reporting "hundreds" of experiences with psychedelics or MDMA. The quality of the eight qualitative studies was high, though one reported no adverse events.

Elevated blood pressure in the psilocybin group was the most common physiological effect, and it was significant in three studies that performed statistical tests. Increased heart rate in the psilocybin group was significant in one study but not in two others. None of the changes in vital signs required medical intervention. In one study, self-limiting severe high blood pressure (systolic BP over 180 or diastolic BP over 110 mmHg) occurred in 4 of 18 patients, which seemed linked to severe anxiety and resolved after therapists offered reassurance. In another study, one patient's blood pressure exceeded protocol limits (diastolic BP over 100 mmHg) but resolved without intervention.

Psilocybin Adverse Reactions

For Depressive Disorders: A small open-label study on psilocybin for treatment-resistant depression reported anxiety, confusion, and nausea as the most common immediate adverse events. Anxiety usually occurred before or at the start of sessions, while confusion happened during the peak of the session. Severity was mild to moderate, except for one case of severe anxiety during a high-dose session. Both MDD studies reported headaches. In one study, patients reported moderate to strong emotional and psychological adverse events, describing feelings like crying, sadness, emotional and/or physical suffering, grief, and isolation. One participant experienced confusing traumatic memories, which worsened their depression for several weeks after treatment. Another patient became unable to communicate during the psilocybin session.

Two qualitative studies illustrated patient experiences: one patient worried about the music creating a melancholic feeling, describing "massive resistance" and a "fearful response" to sensory input, afraid to open eyes and feeling the music was the last thing they would hear. Another patient described a lot of "really, really deep sadness: the loss the grief, it was love and sadness together, and letting go," finding it a process of "unblocking."

Post-treatment headache was the only reported late adverse event; the open-label study only reported headaches in the high-dose (25 mg) group. In one study, 67% rated one of the psilocybin sessions among their top five most psychologically challenging experiences, with 30% calling it the single most challenging. This question was not asked in other studies.

For End-of-Life Anxiety: Two studies on end-of-life anxiety reported immediate and temporary adverse events in the psilocybin groups, including physical discomfort (headache, nausea) and psychological discomfort (anxiety, temporary thought problems, and suicidal thoughts). One study did not report on adverse events. Only one study reported headache as a late adverse event. One patient died by suicide 11 days after a low, placebo-like dose (1 mg) psilocybin session. The patient, who reported feeling bored and wanted to leave the session early, was subsequently removed from the study. The authors concluded that this serious event was not related to research procedures or psilocybin.

Three qualitative studies described patient experiences. One patient reported the experience hit them "very strong" and was "terrifying," feeling "completely disoriented" and "lost in space." They remembered the therapists were there, reached out, and were told, "It’s all right. Just go with it. Go with it," which they did.

For Substance Use Disorders: Two open-label studies investigated psilocybin as a treatment for substance use disorder. One study reported strong (5 of 15 patients) to extreme (1 of 15 patients) anxiety and feeling trapped as immediate adverse events, and headache as a late adverse event. The other study reported few adverse events, though qualitative studies showed how frightening the experience was for some. One patient described the whole experience as "a fever nightmare." Another patient in a second session, after a higher dose, experienced an "amplification of thought" leading to a "confused and chaotic state." Beneath this, they found "a deep well of overwhelming sadness," but eventually surrendered control and entered a state of peacefulness until their thoughts completely quieted.

For Migraines: The most commonly reported immediate adverse events in the migraine study were nausea, anxiety, and lightheadedness. Headache and migraine attacks were the only late adverse events. All were more common in the psilocybin group.

For Demoralization: The most commonly reported immediate adverse events in the open-label study on demoralization in gay men were anxiety, nausea, and headache. Most of these patients (14 of 18) had temporary adverse events, and nearly half (7 of 18) reported them as moderate to severe. Two unexpected late adverse events occurred: one patient experienced a post-traumatic flashback, preventing them from working for two days. Another participant experienced a severe worsening of anxiety, followed by a relapse in methamphetamine use, leading to withdrawal from the study. Despite the severity of these complaints, they were not reported as serious adverse events, which risks underreporting potential negative consequences. However, the longer the time after treatment, the less likely events are still related to the specific treatment.

Ayahuasca Research Findings

Three studies involving 48 patients were included: one randomized controlled trial (RCT) in patients with social anxiety disorder (SAD), one RCT in patients with treatment-resistant depression (TRD), and one open-label study in MDD patients. All involved a single administration of ayahuasca.

None of the studies systematically assessed adverse events, and only one reported immediate adverse events spontaneously mentioned by patients or observed by therapists. At the start of the study, all participants had no prior experience with ayahuasca, and most had no prior experience with other psychedelic drugs.

No significant increases in heart rate or blood pressure were observed in the studies that reported physiological parameters. One study did not report on these.

Ayahuasca Adverse Reactions

Nausea, vomiting, diarrhea, and anxiety were more common immediate adverse events in the ayahuasca group. In one study, anxiety and headache were more common in the control group. In the TRD study, four patients remained hospitalized for seven days due to a "more delicate condition." In the SAD study, one participant experienced an intense episode of temporary fear of dying or going crazy, distress, and feeling disconnected, which resolved after therapists provided reassurance. In this study, no adverse events were reported in the control group. The open-label study only reported vomiting as an immediate adverse event. No studies reported on late adverse events.

LSD Research Findings

Four studies involving a total of 27 patients were included. This comprised one RCT with a nested qualitative study describing end-of-life anxiety treatment, one observational study on compassionate use of LSD and/or MDMA in group therapy (mostly with PTSD and/or MDD patients), and one case report on LSD treatment for a complex personality disorder. Active doses of 100–200 μg were used.

Adverse events were collected through spontaneous patient reports and therapist observations. In the compassionate group therapy study, 39% of participants had prior drug (including cannabis) experience. In the RCT, only one participant (8%) had prior drug experience. The qualitative study was of medium to high quality. Adverse event severity was reported only in the RCT.

No significant differences in heart rate or blood pressure were observed between groups.

LSD Adverse Reactions

Immediate, temporary adverse events included illusions, feeling cold, feeling abnormal, and anxiety. All of these, except moderate-to-severe anxiety, were more common in the high (active) dose groups. Mild-to-moderate emotional distress was equally severe in both low and high-dose groups. Illusions, feeling abnormal, and feeling cold were also late adverse events for one or two patients in the high-dose group. Severity was mostly mild to moderate.

One patient described emotional distress and anxiety, and later resolution, as follows: "The first trip was a panic trip. With almost pure fear of death. It was agony... Really, I had the feeling ‘that I am dying’. Yes, it was just really black, the black side. I was afraid, shaking... It was total exhaustion, not seeing an exit, no escape. It seemed to me like an endless marathon... that was a big part of the trip until it finally led to relaxation... During the second trip, the dark side also showed up at the beginning, but for a rather short time. I was a little tensed, sweating, but not for so long and suddenly a phase of relaxation came. Completely detached. It became bright. Everything was light. It became a pleasant feeling, a warm feeling. No pain. Almost a little floating, clear, being carried and together with the music... It was really gorgeous... The key experience is when you get from dark to light, from tension to total relaxation."

Discussion

This is the first review of adverse events connected to the medical use of MDMA and classic psychedelic substances. By combining quantitative and qualitative studies, it shows that MDMA and psychedelics seem to be generally well-tolerated in treating various mental disorders. However, caution is advised because this conclusion is based on a limited number of relatively small studies, which varied in their design quality and the methods used to assess adverse events. They also used different types and doses of psychedelics in patients with different disorders, and many patients had prior experience with psychedelic drugs.

Acute Adverse Events

The most common immediate physical adverse events in the MDMA groups were fatigue, lack of appetite, feeling cold, thirst, jaw clenching, and sweating. Anxiety and difficulty concentrating were the most common immediate psychological adverse events. For psilocybin studies, the most common immediate adverse events were moderate to severe anxiety, headache, and nausea. Immediate psychological adverse events for psilocybin and LSD (e.g., paranoid thoughts, feeling trapped, illusions, feeling abnormal, psychological discomfort) mostly resolved during the sessions, but were sometimes severe. In one ayahuasca study, a participant temporarily felt disconnected, fearing they were dying or going crazy. Therapists' calming interventions helped resolve this episode, highlighting the importance of trained, skilled, and trusting therapists. Therapist training is also important to help them understand the context of some adverse events and how to effectively prevent or manage them.

Multiple qualitative studies described experiences that were occasionally terrifying, frightening, and confusing. In many cases, patients later viewed these experiences as therapeutically beneficial. Nausea and vomiting were common immediate adverse events in ayahuasca studies. Except for a single case of irregular heartbeats in one MDMA-treated patient with PTSD, no drug-related adverse events or serious adverse events required medical intervention.

Many of the physiological adverse events described in this review are already known from studies on healthy individuals, such as nausea and vomiting with ayahuasca, headaches after psilocybin use, and temporary increases in vital signs, jaw clenching, sweating, and lack of appetite with MDMA. Across studies and disorders, immediate psychological adverse events like anxiety, temporary paranoia, confusion, despair, and psychological distress were commonly reported. These ranged from relatively mild to severe in some cases, and most resolved during the session.

Late Adverse Events

The most common late adverse events across MDMA studies were fatigue, headache, anxiety, difficulty concentrating, and low mood. A small number of patients reported anxiety, depression, panic, and suicidal thoughts in the days and weeks following treatment sessions. Low mood was a late adverse event present in both MDMA and placebo groups. Low mood is also commonly reported three to four days after MDMA use in studies with healthy individuals and by recreational users. However, some studies do not report post-acute decreases in mood. Patients in control groups may have experienced low mood due to a lack of symptom relief and related disappointment. Anxiety in the MDMA groups might be related to re-experiencing traumatic memories and painful emotions. Future research should more accurately describe anxious states and low mood after MDMA treatment and determine if they are a temporary part of the therapeutic process or a subacute adverse effect.

One ayahuasca study reported hospitalizing four patients for a week, without further explanation of how their 'delicate condition' related to the effects of ayahuasca. Two studies reported unexpected adverse reactions: a post-traumatic flashback, a severe worsening of anxiety, and drug use relapse, which occurred in three patients several weeks after psilocybin administration. These patients' conditions worsened and required further treatment; one withdrew from the study. The authors noted that these reactions and the high incidence of anxiety weeks after the session might be related to the patients' complex medical conditions and high rates of other mental health issues. Despite the clinical severity of these complaints, they were not reported as serious adverse events, which risks underreporting potential negative consequences. On the other hand, the longer the time after treatment sessions, the less likely it is that events are still related to a specific treatment. Since individual responses and adverse reactions that may develop weeks or months after treatment cannot currently be predicted, it is crucial for researchers and clinicians to conduct long-term follow-ups and report any undesirable reactions.

Recently, several cases were reported where study participants in an MDMA trial stated they became suicidal after the trial ended, including one case where a participant became overly attached to a therapist and was subsequently abused. Such serious violations occur in the field and must be forcefully addressed. The field of psychedelic treatments might be particularly at risk, as patients under the influence are vulnerable, easily influenced, and highly dependent on the therapist(s). Furthermore, therapists may develop their own personal approaches to therapy that do not always follow professional rules of conduct. This also highlights the importance of thoroughly and systematically monitoring adverse effects.

Suicidal Thoughts

One study reported a suicide 11 days after a session with a placebo-like dose (1 mg) of psilocybin, which was deemed unrelated to study participation. This review did not include the results of a recently completed study comparing very low (1 mg), medium (10 mg), and high (25 mg) doses of psilocybin in patients with treatment-resistant depression (n=233). Overall adverse event rates in that study seem roughly comparable to those summarized in this review. A preliminary publication reported suicidal thoughts, suicidal behavior, and/or intentional self-injury in 7 of 79 patients (8.8%) in the 25 mg group, 6 of 75 patients (8%) in the 10 mg group, and 4 of 79 patients (5%) in the 1 mg group. Fourteen instances in nine patients were reported as serious adverse events. These numbers suggest higher rates of suicidal thoughts in active dose groups (10 and 25 mg), but further interpretation can only be made once peer-reviewed results are published. Suicidal thoughts and behavior were reported in three MDMA studies, though the incidence was low, rates were lower or similar in MDMA groups compared to controls, and serious adverse events were only reported in control groups. Nevertheless, careful assessment of treatment-related consequences is necessary. Being assigned to the inactive group, combined with despair and overly high expectations in patients with severe disorders, may contribute to feelings of hopelessness and subsequent suicidal thoughts or behavior. It is important to remember that mental disorders generally increase the risk of suicidal thoughts. Future studies need to carefully assess suicidal ideation and explore ways to best support all patients long-term.

Understanding Difficult Experiences

Psychological or emotional "adverse events" can be unclear, subjective, context-dependent, and open to interpretation. The qualitative studies in this review showed how patients, in hindsight, considered working through difficult feelings and emotions to be therapeutically beneficial. This is supported by other studies showing that users often transform challenging experiences into personally, morally, or spiritually meaningful ones. The shift in language—using "challenging experiences" instead of "bad trips" (adverse events)—indicates a growing acceptance that difficult experiences are a natural part of psychedelic treatments and can be both distressing and therapeutically beneficial. More narrative evidence and the systematic use of specialized questionnaires can help interpret such experiences and their impact on treatment outcomes. This raises some important questions.

First, because the effects of psychedelics are greatly influenced by context, this may also apply to adverse effects. Treatment designs that reduce or minimize positive contextual factors (such as time spent preparing patients, number of therapists, strength of the therapeutic relationship, time spent providing aftercare and integration) may increase the occurrence of adverse events. Second, some definitions of "adverse event" are not well-suited for psychedelic treatment. By definition, adverse events are undesired and should be avoided. Categorizing challenging but potentially beneficial experiences as adverse events can therefore be counterproductive. For instance, labeling "anxiety" as an adverse event might prompt patients to try to avoid difficult feelings and thoughts instead of engaging with them as part of the healing process. Distinguishing between truly adverse elements that should not be part of the therapeutic process (e.g., physical adverse events like headaches and nausea, or worsening of distressing symptoms like suicidal thoughts and dissociation) and psychological experiences that are part of the therapeutic process is crucial to better understand how psychedelic treatments work. To illustrate this complexity: vomiting, the most commonly reported adverse event in ayahuasca studies, is often considered a therapeutic element in both traditional indigenous and modern ayahuasca practices. In the case of anxiety, the severity and length of the difficult episode may also predict less than optimal or even negative clinical outcomes. Careful screening, preparation, session monitoring, psychological support, and post-session integration are essential for maintaining a low incidence of adverse events and maximizing treatment effectiveness. Without these, young people and those with certain personality traits (like neuroticism, emotional instability, or low openness to experience and acceptance), as well as those in an apprehensive, preoccupied, or confused mental state, may be more prone to challenging experiences that could worsen their existing problems. Finally, recent articles have described how psychedelics may change a person's fundamental beliefs about reality. Such experiences can be very meaningful but can also cause an "ontological shock," especially when they conflict with prior religious, personal, or spiritual beliefs. This introduces ethical and metaphysical challenges, and a deep responsibility for therapists to reduce potential difficulties when patients face profound metaphysical experiences.

Study Limitations

This review has both strengths and limitations. An important strength is the systematic collection of studies and the combination of quantitative and qualitative research. However, a key limitation is that the included studies were mostly small, varied in design, drug dose, and patient demographics, making comparisons difficult and preventing a combined statistical analysis (meta-analysis). Second, few studies assessed adverse events consistently or systematically. This is a common issue in psychopharmacology and not unique to psychedelic drug studies. Instead, most studies relied on adverse events reported spontaneously by patients and/or observed by therapists. Spontaneous reporting depends on individual and contextual factors that allow patients to recall or discuss negative experiences. Therapist observation is similarly unclear, as observations may be influenced by their subjective perception and interpretation. Five pilot studies did not report any adverse events. While it is possible, though unlikely, that no adverse events occurred, not reporting any adverse events is similar to significant omissions in early psychedelic research, where researchers often did not fully disclose the frequency and severity of negative reactions. Third, there are not enough placebo-controlled trials to draw strong conclusions about the direct cause-and-effect relationship between psychedelic treatment and adverse events. It is especially difficult to truly blind study participants to their treatment condition in psychedelic drug studies, particularly given the high percentage of patients who have prior experience with psychedelics. Also, due to the small sample sizes, studies may have missed less frequent adverse events, hindering broader application of findings, as some adverse events may only become apparent in larger populations. All studies in this review refer to a total of 598 unique patients, with 521 receiving an active dose. Fourth, the severity and timing of adverse events (especially for late ones) were reported in only a few studies, making it difficult to properly interpret the reported adverse events.

Next, most trials had strict participation criteria, involving extensive screening and excluding patients with almost any co-occurring mental disorder. This included a personal or family history of psychotic, bipolar, dissociative identity, eating, and substance use disorders, and active suicidal thoughts—all of which are common in patients with more severe mental disorders. Sixth, many studies included patients with previous experience with the experimental drug; up to 56% for MDMA and about half of all psilocybin study participants had prior experience with psilocybin or other psychedelics. The intensity of drug effects and experienced difficulties seem to be lower and less severe in those with past experience with these drugs, which limits how broadly the findings can be applied to wider patient populations. Additionally, minority groups are underrepresented in clinical psychedelic research. Therefore, it is possible that patients new to psychedelics, with different social and cultural backgrounds and more severe mental health conditions, may have more intense experiences or greater difficulty coping with potentially disruptive events. That said, the only study in this review that compared lifetime psychedelic use and the degree of experienced challenge within the session found no significant correlations, suggesting that previous experience does not necessarily predict the occurrence of adverse events in clinical populations or contexts. Seventh, while qualitative studies helped provide context for some of the immediate challenging experiences, they were not always specifically designed to explore such effects. Furthermore, qualitative studies are unevenly distributed and relatively scarce; there were none on ayahuasca, only one each on MDMA and LSD, and eight on clinical psilocybin studies. Finally, any conclusion on the frequency and interpretation of adverse events should consider the patient's individual state ("set") and the treatment environment ("setting"). More information about "set and setting" should be provided when reporting and interpreting the results of clinical trials involving psychedelics.

Conclusion

This first review of immediate and late adverse events in treatments with psychedelic substances and MDMA, combining both quantitative and qualitative studies, shows that all compounds immediately cause temporary headaches, nausea, and anxiety. Anxiety covers a wide range of psychologically challenging experiences that might also be helpful for the therapeutic process. Separating truly adverse events from potentially beneficial effects is complex but crucial to improve understanding of psychedelic treatments. Serious adverse events appear to be largely absent based on the included studies, as were lasting physiological side effects. Suicidal thoughts and behavior were rare but did occur; these represent a serious psychiatric emergency, which highlights the importance of investigating which patients are most at risk and how to best reduce the likelihood of these occurrences. Overall, adverse events were inconsistently defined and not adequately assessed. Future studies should describe the timing and severity of effects more thoroughly, perhaps using scales specifically designed for challenging experiences. Many studies included patients with prior psychedelic experience, which might bias results and limit how broadly they can be applied. Qualitative research can also add nuance by detailing and explaining the meaning of challenging experiences. Full transparency about adverse events is a responsibility of healthcare providers, especially in a new field driven by enthusiastic researchers. Understanding the full range of unpleasant, potentially harmful, and transformative treatment-related events is crucial to prepare future therapists who might otherwise be insufficiently ready to handle challenging and potentially destabilizing patient experiences, particularly in larger groups of patients with more complex and possibly co-occurring conditions.

Introduction

Research into how psychedelic drugs affect mental health has shown some good results for different mind problems. However, many of these studies were small and only looked at certain people. Now, bigger studies are finishing, and new uses for these drugs are being explored. This means some of these drugs might soon be approved for use.

Using psychedelics without proper medical supervision can cause bad effects. These include feeling worried, panic, sadness, paranoia, or acting in unsafe ways. Such events can sometimes lead to lasting mental problems. Even though most bad effects happened when people used these drugs outside of medical settings, and even if small medical studies have shown them to be safe, giving them to more kinds of patients with other health issues might bring unexpected problems. Also, bad reactions to medicine can make people stop their treatment. It is important to find out all the possible bad effects of psychedelic drugs, especially for people with mental problems that are hard to treat. No full review of this topic has been done since 1984. Earlier reviews, from 1960 and 1984, suggested that bad effects were rare, often mild, and usually happened with people who already had certain mental conditions. Some people reported "flashbacks," which are like reliving parts of the drug experience. These are now called Hallucinogen Persisting Perception Disorder (HPPD) and are usually mild.

It is hard to check for bad effects for many reasons. Sometimes, these effects are not specifically looked for. There are many different types of reactions, and how they are reported can be messy. Also, different words are used to mean the same thing, which makes it confusing. The way a psychedelic drug affects a person can also change a lot based on their feelings and the setting.

Since more patients with mental problems might be treated with psychedelics soon, a full picture of both the good and bad effects is needed. This will help doctors decide if the benefits are worth the risks and understand which patients might benefit most. This paper brings together information about any bad effects that happened during or after treatments with common psychedelics like psilocybin, LSD, ayahuasca, and MDMA in patients.

Methods

This study looked at different research papers in an open way to find out about bad effects, rather than trying to prove one specific idea. The review did not set rules beforehand. The study followed guidelines for reviewing medical papers, pulling out information from both number-based and experience-based studies. A "bad event" was defined as any unwanted or harmful event that started during or after taking a classic psychedelic or MDMA, whether or not it was caused by the drug. "Immediate bad events" happened during or on the day of the drug session. "Later bad events" happened after the drug session day. Other terms like "treatment-emergent bad events" were not used in this study to keep things simple.

Selection criteria

All medical studies from 2000 onward that described treating patients with mental problems using classic psychedelics (like psilocybin, LSD) or MDMA were included. This included different kinds of studies, even single patient stories. Studies on healthy people, surveys, or reviews were not included. The goal was to find any possible bad reactions in patients with a mental health diagnosis who received these drugs. No specific rules were made to keep out any findings, so that the study could be as broad as possible and not miss any unwanted results.

Search methods

Researchers looked through medical and psychology databases in July 2021. They used keywords for psychedelic drugs and study types, but did not include animal studies. They also checked a website for ongoing medical studies. In addition, they looked through lists of references in the papers they found.

Quality assessment

A formal check of study quality was not seen as most important because even high-quality studies might not report bad effects well. Instead, the study looked at how bad effects were checked or reported, who was included or not included in the study, and how many patients had used the drug before. A specific checklist was used for experience-based studies.

Data extraction and synthesis

All facts and numbers about immediate and later bad effects, when they were checked, and any very serious bad events were taken from the papers. The number of patients who had a bad effect was rewritten as a percentage. For studies that collected patient experiences, ideas and descriptions about bad effects were looked for. These were then used to help explain and add to the findings from the number-based studies.

Results

The study started by finding 5640 articles. After removing repeats, 3709 titles and summaries were checked. After careful checking, 61 full articles were looked at, and 44 of them were included. Also, 596 registered medical trials were found, and 24 completed ones were included.

The included studies were published between 2006 and 2021 and looked at a total of 598 different patients. Of these, 521 received the active drug. The number of patients in each study was small, from 1 to 105. The average age of patients was between 25 and 59 years. If a study had more than one paper about it, these papers were combined. All studies about patient experiences were based on parts of the number-based studies that were included. The studies used different designs, drugs, doses, and looked at different problems. Four studies with numbers and two with patient experiences did not report on any bad effects.

This section will describe the main results for each drug, focusing on the most common bad effects.

MDMA

Sixteen studies were included for MDMA, with a total of 266 patients. Most of these studies (10 papers) looked at MDMA for Post-Traumatic Stress Disorder (PTSD). Other studies looked at it for anxiety at the end of life, social anxiety in adults with autism, tinnitus (ringing in the ears), and alcohol problems. Patients usually received doses from 50 to 125 mg.

Many MDMA studies only reported bad effects if patients mentioned them on their own. Some patients in the studies had used MDMA or "ecstasy" before, ranging from none to more than half.

During MDMA sessions, patients often had slightly higher blood pressure, body temperature, and heart rate. These changes were not serious and did not need medical help. One serious bad event related to MDMA was reported where a patient's heart rhythm changed, needing a night in the hospital, but it got better on its own.

For patients with PTSD, the most common immediate physical bad effects were jaw clenching or tight muscles, headaches, feeling sick to the stomach, tiredness, and not wanting to eat. Feeling worried was the most common mental bad effect. Most immediate bad effects happened more often with MDMA than with a fake pill, but they were mostly mild to medium. The amount of MDMA given did not seem to change how many bad effects happened. In two studies, very serious bad events were reported. In one, two patients had thoughts of self-harm, but these were in the group that did not get MDMA. In another study, three of four serious events were not thought to be related to MDMA. One patient's heart problem got worse, but it was likely related to MDMA. One patient's story showed they felt strong feelings and physical reactions like muscle tightening while thinking about past traumas.

Common later bad effects were tiredness, not wanting to eat, low mood, trouble sleeping, needing more sleep, feeling easily annoyed, headaches, trouble focusing, and feeling worried. Low mood was reported in both the MDMA and fake pill groups. Some patients reported ongoing worry, low mood, or panic attacks up to 2 months after their last session.

For patients with end-of-life anxiety, immediate bad effects were jaw clenching, thirst, dry mouth, sweating, and headaches. Later bad effects included tiredness, needing more sleep, trouble sleeping, feeling worried, jaw clenching, and low mood. Most of these happened more in the MDMA group.

For social anxiety in adults with autism, more immediate bad effects happened in the MDMA group, such as feeling worried, trouble focusing, tiredness, headaches, and feeling cold. Common later bad effects included tiredness, headaches, trouble focusing, low mood, and needing more sleep. Most of these were more common in the MDMA group. Some patients in both the MDMA and fake pill groups had low mood, thoughts of self-harm, and panic attacks after the first drug session.

Studies on tinnitus and alcohol use disorder did not report on bad effects.

Psilocybin

Twenty papers were included for psilocybin, looking at 257 patients. These included different kinds of studies for anxiety at the end of life, major depression, migraine headaches, obsessive-compulsive disorder, hard-to-treat depression, alcohol use disorder, tobacco smoking, and low spirits in older AIDS survivors. Doses varied.

Two psilocybin studies did not report bad effects. Most others relied on what patients said or what therapists saw. Some also used interviews or surveys. Three studies reported how serious the bad effects were. In studies that looked at different doses, one found that some mental bad effects were more common with a medium dose than a high dose. Another found more headaches after treatment with a higher dose.

Most studies asked if patients had used psychedelics before. About half of all psilocybin study patients had used psilocybin or other psychedelics before the study.

Patients taking psilocybin often had slightly higher blood pressure, which was clear in three studies. Heart rate also went up in some studies but not others. No changes in vital signs needed medical help. In one study, four patients had very high blood pressure that went down on its own after they were calmed by therapists.

For patients with depression, immediate bad effects were feeling worried, confused, and sick to the stomach. Worry often happened before or at the start of the session, and confusion happened when the drug was at its strongest. Most bad effects were mild to medium, but one case of severe worry happened. Both depression studies reported headaches. In one study, patients felt strong emotional pain, sadness, and loneliness. One patient had confusing, upsetting memories which made their depression worse for weeks. Another patient could not talk during the session. Some patients said the music made them feel sad and afraid, or that they felt a lot of deep sadness, loss, and grief, but that it helped them let go.

Headaches after treatment were the only later bad effects reported. In one study, many patients (67%) said a psilocybin session was one of the top five most mentally challenging experiences they ever had.

For end-of-life anxiety, immediate and short-term bad effects in the psilocybin groups included physical discomfort (headaches, feeling sick) and mental discomfort (worry, strange thoughts, and thoughts of self-harm). One study did not report bad effects. Only one study reported headaches as a later bad effect. One patient died by suicide 11 days after a very low dose session. The study authors said this was not related to the research or the drug. One patient described a very strong and terrifying experience where they felt lost and scared, but then felt better when a therapist held their hand and told them to "go with it."

For substance use disorders, one study reported strong to extreme worry and feeling trapped as immediate bad effects, and headaches as a later bad effect. Another study reported few bad effects, though patients described the experience as a "fever nightmare" or a "confused and chaotic state" followed by deep sadness, but eventually leading to peace.

For migraines, the most reported immediate bad effects were feeling sick, worried, and dizzy. Headaches and migraine attacks were the only later bad effects, and they were more common in the psilocybin group.

For low spirits in gay men, the most common immediate bad effects were worry, feeling sick, and headaches. Most of these were short-term, and nearly half were moderate to severe. Two unexpected later bad effects happened: one patient had a flashback that kept them from working for 2 days. Another patient had much worse worry, which led to using methamphetamine again, and they left the study.

Ayahuasca

Three studies were included for ayahuasca, with 48 patients. These included studies for social anxiety disorder and hard-to-treat depression. All studies involved taking ayahuasca one time.

None of the studies systematically checked for bad effects. Only one reported immediate bad effects that patients mentioned or therapists saw. All patients in these studies had never taken ayahuasca before, and most had not taken other psychedelics either.

No big changes in heart rate or blood pressure were seen.

Feeling sick, throwing up, diarrhea, and feeling worried were more common immediate bad effects in the ayahuasca group. In the hard-to-treat depression study, four patients stayed in the hospital for a week due to a "more delicate condition." In the social anxiety study, one patient had a strong, short episode of fear of dying or going crazy and felt disconnected, but felt better after therapists calmed them. This study reported no bad effects in the fake pill group. The other study only reported throwing up as an immediate bad effect. No studies reported later bad effects.

LSD

Four studies were included for LSD, with 27 patients. These included a study for end-of-life anxiety, an observational study on using LSD and/or MDMA in group therapy for PTSD and depression, and a single case report for a complex personality disorder. Doses were usually 100–200 μg.

Bad effects were collected when patients mentioned them or therapists saw them. Some patients had used drugs before. How serious the bad effects were was only reported in one study.

No big differences in heart rate or blood pressure were seen between groups.

Immediate, short-term bad effects included seeing things that were not there, feeling cold, feeling strange, and feeling worried. Most of these were more common in the high-dose groups, except for moderate-to-severe worry, which was similar in both groups. Mild-to-moderate emotional upset was also similar in both groups. Seeing things, feeling strange, and feeling cold were also later bad effects for one or two patients in the high-dose group. Most bad effects were mild to moderate. One patient described their first trip as "pure fear of death," feeling "agony," and "total exhaustion," but eventually it led to relaxation and a "pleasant feeling, a warm feeling" where "everything was light."

Discussion

This is the first review of bad effects related to the medical use of MDMA and classic psychedelics. It brings together studies that use numbers and patient experiences. Based on this review, MDMA and psychedelics seem to be well-tolerated when used to treat different mental problems. However, this idea comes from a small number of studies that varied in quality and design. They used different psychedelics at different doses in patients with different problems, and many patients had used psychedelics before.

Acute AEs

The most common immediate physical bad effects for MDMA were tiredness, not wanting to eat, feeling cold, thirst, jaw clenching, and sweating. Feeling worried and having trouble focusing were the most common immediate mental bad effects. For psilocybin, the most common immediate bad effects were moderate to severe worry, headaches, and feeling sick. Immediate mental bad effects for psilocybin and LSD (like strange thoughts, feeling trapped, seeing things, feeling strange, and mental upset) usually got better during the sessions, but were sometimes serious. In one ayahuasca study, a patient temporarily felt disconnected and feared dying or going crazy. Therapists helped calm this patient down, showing how important trained and trusted therapists are.

Many patient experience studies described trips as sometimes terrifying, frightening, and confusing. But in many cases, looking back, patients felt these experiences were helpful for their treatment. Feeling sick and throwing up were common immediate bad effects in ayahuasca studies. Except for one heart problem in an MDMA-treated patient, no drug-related bad events needed medical help.

Many of the physical bad effects described are already known from studies with healthy people. For example, feeling sick and throwing up with ayahuasca, headaches after psilocybin, and short-term higher vital signs, jaw clenching, sweating, and not wanting to eat with MDMA. Across studies and problems, common immediate mental bad effects included worry, short-term paranoia, confusion, sadness, and mental upset. These were often reported, from mild to severe, and most got better during the session.

Late AEs

The most common later bad effects across MDMA studies were tiredness, headaches, worry, trouble focusing, and low mood. A small number of patients reported worry, depression, panic, and thoughts of self-harm in the days and weeks after treatment. Low mood was a later bad effect seen in both MDMA and fake pill groups. Low mood is also often reported a few days after MDMA use in healthy people or those who use "ecstasy" for fun. However, some studies do not report this. Patients in fake pill groups might have felt low mood because their symptoms did not get better, and they were disappointed. Worry in the MDMA groups might be linked to re-living bad memories and painful feelings. Future research should look more closely at worry and low mood after MDMA treatment to see if they are a short part of the healing process or a real bad effect.

One ayahuasca study reported four patients staying in the hospital for a week, but did not explain how their "delicate condition" was related to ayahuasca's effects. Two studies reported unexpected bad reactions (a flashback after a traumatic event; severe worsening of worry and drug use again; worsening of depression) in three patients that happened several weeks after psilocybin treatment. These patients got worse and needed more therapy; one patient left the study. The authors noted that these reactions, and the high rate of worry weeks after the session, might be because these patients had complex medical problems and many mental health issues. Even though these problems were serious, they were not reported as "very serious bad events," which means potential negative outcomes might be missed. On the other hand, the longer the time after treatment, the less likely events are related to that treatment. Since we cannot predict individual reactions and bad effects that might appear weeks or months later, it is very important for researchers and doctors to do long-term checks and report any bad reactions.

Recently, cases were reported where patients in an MDMA study said they had thoughts of self-harm after the study finished. This included one case where a patient became too close to a therapist and was later mistreated. Unfortunately, such very serious wrongdoings happen in this field and must be strongly addressed. It might be that psychedelic treatments carry a special risk, as patients are vulnerable and easily influenced during the session and are very dependent on the therapist. Also, therapists might have their own ways of working that do not always follow professional rules. This also shows how important it is to carefully and regularly check for bad effects.

Suicidality

One study reported a patient's death by suicide 11 days after a session with a very low dose of psilocybin, but it was said not to be related to the study. A recent, large study on psilocybin for hard-to-treat depression found thoughts of self-harm, self-harm actions, or intentional self-injury in some patients in all dose groups, with more in the active drug groups. Some of these were reported as very serious bad events. These numbers suggest more self-harm thoughts in active dose groups, but more will be known when the full results are published. Thoughts of self-harm were reported in three MDMA studies, but they were rare, and less or equally common in the MDMA groups compared to control groups. However, doctors must carefully check for self-harm thoughts and find ways to best support all patients long-term.

Interpreting adversity

Mental or emotional "bad effects" can be unclear, depend on the person, the situation, and how they are understood. The patient experience studies in this review showed that, looking back, patients often found it helpful to work through difficult feelings and emotions. Other studies also show that people often turn challenging experiences into meaningful ones. The change in words from "bad trips" to "challenging experiences" shows that tough experiences are now seen as a normal part of psychedelic treatments and can be both upsetting and helpful. This brings up some important points.

First, since the effects of psychedelics are greatly influenced by the setting, this might also apply to bad effects. Treatment designs that offer less support (like less time preparing patients, fewer therapists, weaker connection with therapists, less aftercare) might lead to more bad effects. Second, some definitions of "bad effect" do not fit well with psychedelic treatment. By definition, bad effects are unwanted and should be avoided. Calling challenging but possibly helpful experiences "bad effects" can make things worse. For example, if "worry" is seen as a bad thing, patients might try to avoid difficult feelings instead of working through them as part of therapy. It is very important to separate truly bad things (like headaches, feeling sick, or worsening of serious symptoms like thoughts of self-harm) from mental experiences that are part of the healing process. For example, throwing up is a common bad effect in ayahuasca studies, but it is often seen as a helpful part of healing in traditional practices. Also, with worry, how serious and long the difficult episode is can predict bad results. Careful checking, preparation, watching during the session, mental support, and after-session talks are key to keeping bad effects low and making treatment work best. Without these, young people and those with certain personality traits might be more likely to have challenging experiences that could make their existing problems worse. Finally, recent papers have shown that psychedelics can change deeply held beliefs about life and the world. Such experiences can be very meaningful but can also cause a big shock, especially if they go against what someone believed before. This creates ethical and spiritual challenges, and therapists have a deep responsibility to help patients deal with such strong experiences.

Limitations

This review has both strong points and weak points. A strong point is that it systematically collected studies and combined both number-based and patient experience studies. However, there are also important weaknesses. First, the included studies were mostly small and varied a lot in how they were designed, the type and dose of drug used, and the types of patients, making it hard to compare them. Second, only a few studies checked for bad effects in a consistent or regular way. Most studies relied on what patients said on their own or what therapists saw. It is hard to know how accurate these reports are, as they depend on many factors. Five small studies reported no bad effects at all. While this is possible, it reminds us of early psychedelic research where negative reactions were often not fully shared. Third, there were not enough studies comparing drugs to fake pills to say for sure that bad effects were caused by the psychedelic treatment. It is also hard to hide from patients whether they got the real drug, especially since many patients had used psychedelics before. Also, because studies were small, they might have missed bad effects that happen less often. All studies in this review looked at a total of 598 different patients, with 521 receiving an active dose. Fourth, how serious bad effects were, and when they started (especially later ones), were only reported in a few studies, making it hard to truly understand them. Also, most studies had very strict rules for who could join, keeping out patients with almost any other mental problem. This means patients in the studies might not be like most patients with more severe mental problems. Sixth, many studies included patients who had used the drug before. This might make the bad effects seem less serious, limiting how much we can apply these results to all patients. Also, minority groups are not well represented in psychedelic research. It is possible that patients new to psychedelics, from different backgrounds, or with more severe mental health issues might have more intense experiences or have more trouble dealing with upsetting events. However, one study found that past psychedelic use did not predict bad effects in medical settings. Seventh, while patient experience studies helped explain some immediate challenging experiences, they were not always set up to look at these effects specifically. Also, there are not many of these studies, and none for ayahuasca. Finally, any ideas about how often bad effects happen and what they mean should consider the patient's mindset and the setting of the treatment. More information about these "set and setting" factors should be given when reporting psychedelic treatment study results.

Conclusions

This first review of immediate and later bad effects in treatments with psychedelics and MDMA, combining number-based and patient experience studies, shows that all drugs quickly cause short-term headaches, feeling sick, and worry. Worry covers many mentally challenging events that can also be helpful for healing. It is complex but very important to tell the difference between truly bad effects and those that might be beneficial for treatment. Very serious bad events seemed to be mostly absent in the included studies, as were lasting physical bad effects. Thoughts of self-harm and actions of self-harm were rare but did happen; these are very serious medical emergencies. This shows how important it is to find out which patients are most at risk and how to best lower the chance of these happening. Overall, bad effects were not defined or checked for in a consistent way. Future studies should describe when effects happen and how serious they are in more detail. Many studies included patients who had used psychedelics before, which might skew results and limit how widely they can be applied. Patient experience research can also add detail and help understand the meaning of challenging experiences. Being fully open about bad effects is a duty for doctors, especially in a new field driven by excited researchers. Understanding all the unpleasant, possibly harmful, and life-changing events related to treatment is key to preparing future therapists who might otherwise not be ready to handle challenging and potentially upsetting patient experiences, especially in larger groups of patients with more complex and possibly other health issues.