Crisis Intervention Related to the Use of Psychoactive Substances in Recreational Settings - Evaluating the Kosmicare Project at Boom Festival

Over recent decades psychoactive substance (PAS) use patterns have shifted, with recreational settings becoming a dominant context for use in Portugal and elsewhere. Carvalho and colleagues note that while much recreational PAS use is non-problematic and associated with motives such as spiritual growth or creativity, outdoor festivals and raves also concentrate risk: large numbers of people experiment with diverse substances, adulteration and poly drug use are common, and some attendees experience severe or prolonged psychological distress. Existing literature and practice provide models from harm reduction and crisis intervention, and Grof's psychedelic psychotherapy approach is cited as a relevant framework, but there is little formal evaluation of crisis intervention protocols that target psychological emergencies in recreational settings rather than purely medical emergencies handled in emergency departments. This paper reports evaluation research designed to transform Kosmicare (KC) — a crisis-intervention and harm-reduction programme operating at Boom Festival (Portugal) since 2002 — into an evidence-based model. The study aimed to describe the intervention process, test whether KC reduced crisis symptoms among festival visitors, and explore short- and longer-term visitor outcomes and perceptions. The evaluation uses a mixed-methods, naturalistic approach focused on process and outcome indicators gathered from multiple data sources collected during and after festival editions in 2010 and 2012.

The study used a pretest–posttest programme-evaluation design without a control group, combined with qualitative methods and a follow-up inquiry. Data collection occurred in three waves: during KC 2010 (forms completed by sitters and secretaries capturing arrival, intervention and departure data), one-month post-festival e-mailed team feedback, and an online follow-up questionnaire completed by visitors six to eighteen months after intervention. Quantitative analyses were performed with PASW 18 and qualitative analysis with QSR NVIVO 9. Participants comprised several samples. Sample 1 (intervention cases) included n=122 situations attended in 2010; Sample 2 comprised n=36 team members (about 70% of staff) who returned feedback; and Sample 3 was a follow-up sample of n=18 visitors recruited from the 2010 and 2012 editions (very low return rate from the available e-mail contacts). In total the extracted text reports n=176 participants across samples, but sample sizes vary substantially by measure and many cases lack complete data. KC operates continuously during festival days with a multidisciplinary, largely volunteer team (sitters, team leaders, medical staff, secretaries). Intervention emphasised proximity, non-directive psychotherapeutic support, a safe restful space, basic physiological care, and selective medical backup; pharmacological sedation was discouraged except in limited circumstances. Visitors arrived via self-referral, escorts, or festival staff. Measurement relied principally on observer-rated forms completed by sitters and secretaries. A bespoke observer checklist—the Mental State Examination Checklist (MSEC)—was used to rate mental state on arrival and at discharge; its psychometric evaluation had begun by 2012 with preliminary reports of high internal consistency but remained under development. Other instruments captured intervention processes, team perceptions (SWOT and questionnaires), and follow-up visitor-reported outcomes. The researchers treated process and outcome indicators separately and analysed pre/post symptom change using paired t-tests on available matched data, while qualitative responses and team feedback were analysed thematically.

Participants attended KC for a range of reasons: difficult experiences involving PAS (intentional or unintentional), personal crises not explicitly involving PAS, non-crisis requests (rest, hydration, minor care), and mental health outbursts that sometimes appeared independent of substance use. Demographically, visitors were predominantly male, aged 19–39, European and relatively highly qualified, but detailed demographics were limited because intake interviews were kept brief. Self-report measures indicated frequent poly drug use: over 40% of visitors reported using multiple substances and, according to the abstract, 52% reported LSD use. The most common drug combinations reported were LSD+alcohol (12%), MDMA+cannabis (10%), amphetamines+alcohol (7%), and LSD+MDMA (8%). The researchers emphasise that these figures are based on visitors' beliefs about substances ingested and that adulteration and mislabelling are common, limiting attribution of effects to particular drugs. Symptom patterns recorded via the MSEC (data available for n=83 on symptom profiling, and n=44 for paired pre/post analyses) showed affect- and emotion-related symptoms (anxiety, fear, crying, suicidal ideation) to be the most frequent across substances. LSD and MDMA were associated with larger numbers of symptoms in sitter reports; other substances showed characteristic patterns (e.g. alcohol linked to aggression and affective lability; cannabis to confusion and disorientation). A gender analysis indicated a predominance of crisis symptoms in male visitors, with the mean difference marginally significant. Intervention characteristics: about 50% of episodes were resolved within 1–5 hours, 31% required about two shifts, and 13% lasted over 24 hours up to several days. Sitters predominantly used psychotherapeutic strategies (non-directive talk therapy, active listening, empathy), supplemented by logistical supports (rest area, warm clothing) and, rarely, medical/homeopathic remedies. Physical contact (massage) was used selectively and avoided when paranoia was present. Outcome indicators: paired pre/post comparisons on the MSEC for n=44 visitors yielded a mean symptom difference of 6.84, which was statistically significant (p<.000), interpreted by the researchers as evidence of crisis resolution. Using another process indicator from Form 3 on a valid subsample (n=54), 76% (n=41) of crisis episodes were classified as resolved, 17% (n=9) unresolved, and 7% (n=4) interrupted. Unresolved episodes were disproportionately associated with mental health outbursts that showed persistent symptoms, whether related to PAS use or not. Visitor satisfaction at departure (valid n=58) was high: 81% (n=41) 'totally agreed' with “I have been helped by KC”; 75% 'totally agreed' staff were helpful and caring; and 80% 'totally agreed' staff were well prepared. Team satisfaction was also positive: roughly 80% agreed the project implementation was high, and over 90% considered the intervention well accepted by visitors. Follow-up data (n=18) are limited but largely positive: 15 of 18 reported KC effective in resolving their crisis, 12 considered benefits to have lasting impact, and respondents reported longer-term changes such as increased responsibility towards drug use (n=5), improved knowledge for handling bad trips (n=2), and greater self-knowledge or improved relationships (n=5). A small number reported negative or worsening integration (a few cases of increased symptoms or persistent perceptual changes).

Carvalho and colleagues interpret their findings as evidence that systematic process and outcome indicators can be collected in an unconventional, naturalistic festival setting and that KC appears to have demonstrable impact in resolving many acute crisis episodes and in providing highly acceptable support to a hard-to-reach PAS-using population. The programme's emphasis on a safe, supportive care space and on therapeutic skills such as empathy and active listening is seen as central to its effectiveness, and the predominance of non-pharmacological strategies aligns with the intended crisis-intervention model. The researchers highlight poly drug use patterns and the predominance of LSD and MDMA in reported crisis episodes, noting that these combinations — and the presence of adulterants — complicate attempts to link specific substances causally to particular symptom profiles. They also acknowledge an unresolved debate about whether PAS precipitate mental disorders or merely unmask pre-existing vulnerability; in their data, unresolved crises were more often associated with suspected pre-existing psychopathology, suggesting the need for careful follow-up and further research into trajectories of risk. Key limitations are acknowledged throughout: the naturalistic, non-randomised pretest–posttest design without a control group; substantial missing data and varying sample sizes across measures; low and selective follow-up response rates limiting generalisability; reliance on sitter-observed measures and on visitors' self-reported substance use; and an instrument (MSEC) whose psychometric work was preliminary at the time of reporting. These constraints restrict causal inference and the scope of conclusions about long-term impact. Despite these limits, the authors suggest practical and research implications grounded in their findings: further psychometric validation and refinement of instruments, longitudinal follow-up studies to examine substance-use trajectories and mental health outcomes, strengthened partnerships with external mental health services for cases suspected of pre-existing disorder, and operational changes informed by team feedback (for example relocating the care space, improving security and communications). They present KC as a feasible, acceptable model for proximal crisis intervention and harm-reduction in large-scale recreational settings while calling for more rigorous trials and expanded monitoring to establish robust evidence of effectiveness.