Barriers to help-seeking among music festival attendees in New South Wales, Australia.

INTRODUCTION: Prompt help-seeking behaviour by music festival attendees can reduce risks associated with drug use; however, little is known about perceived barriers to help-seeking when experiencing or witnessing illness at music festivals. We explored potential barriers and their association with festivalgoer characteristics.
METHODS: We conducted an on-site cross-sectional survey of attendees at New South Wales music festivals in 2019/2020. Perceived barriers to help-seeking in the hypothetical event of the respondent or a friend becoming unwell at the festival were assessed, and regression analyses were conducted to identify characteristics associated with these barriers.
RESULTS: Across six festivals, 1229 people were surveyed and four-fifths (83.2%) reported ≥1 barrier: 32.7% fear of getting in trouble with the police, 20.6% not knowing where to find help, 17.2% not knowing how unwell someone might be and 15.3% concern about friends or relatives finding out. In multivariable analyses, people of diverse sexuality and people using drugs that day had greater odds of reporting fear of trouble with the police. People reporting drug use that day had lower odds of reporting not knowing where to find help. Men, gender-diverse people and people using drugs that day had greater odds of reporting concern about friends or relatives finding out. DISCUSSION AND
CONCLUSIONS: Our data substantiate concerns regarding policing strategies and their impact on festivals. Initiatives to support conversations about drugs with friends and families may be best targeted to younger people and those from gender-diverse backgrounds.

Introduction

The use of illicit substances at music festivals is common, with recent Australian reports estimating around a quarter to a third of festival attendees engage in drug use during the event [1, 2]. While the majority of this drug use occurs with minimal health consequences, it can be accompanied by a range of harms including injury, hyperthermia, coma and in rare instances, death [3].

Barriers to help‐seeking at festivals, in the context of drug toxicity, may contribute to delays in presentation for emergency medical treatment and hence more severe toxicity and associated harms. While there is a wealth of information regarding barriers to treatment‐seeking in people with substance use disorders and help‐seeking following a drug overdose in other settings [4, 5, 6], there is a substantial gap in the literature regarding help‐seeking following drug toxicity at music festivals. Findings between these settings may not be directly transferable, but given the similarities in the experience of stigma and barriers to treatment among different populations of people who use drugs, the former provides useful information and background for this exploration of barriers for other populations with highly prevalent drug use. Studies of opioid overdose report up to half of all people witnessing an overdose experiencing barriers and not seeking help [4, 5]. A number of barriers to help‐seeking are frequently reported following overdose; at the forefront is fear of police involvement and legal repercussions [4, 5, 6] and this has been correlated with a propensity to prohibitionist policing policies [7]. Other reported barriers and causes of delays in help‐seeking following opioid overdose included attempting to ‘ride it out’, prior experience with overdose, others present obstructing help‐seeking and taking measures to reverse the overdose, such as taking other countermeasure drugs [5, 6, 8]. In the context of drug‐related emergencies, younger people, men and those with experience of stigma from health professionals were less likely to seek help [5, 6, 8]. Groups previously identified as experiencing barriers to help‐seeking for substance use disorder include younger people and people from diverse gender or sexuality groups [9, 10, 11, 12].

By contrast, there is a dearth of evidence surrounding barriers to help‐seeking at music festivals in the context of drug toxicity [8]. A recent Australian Coronial inquest into deaths at festivals highlighted a number of contributory barriers, including policing strategies and stigma, not knowing how unwell someone was and challenges in talking with family about drug use [13]. Australian and international research details the potential for unintended harmful consequences of law enforcement strategies at festivals, including hazardous drug use behaviour [14, 15, 16]. However, the experience and correlates of barriers to help‐seeking in the context of illness (whether related to drug toxicity or otherwise) at music festivals are poorly understood. It is essential to understand barriers to accessing care in the festival setting and their associations with festivalgoer characteristics to identify populations that may be at particular risk and to design targeted strategies to overcome such barriers.

This study surveyed attendees on‐site at music festivals in New South Wales (NSW), Australia during the 2019–2020 festival season. Our aims were to describe barriers to help‐seeking and their association with key festivalgoer characteristics, including age, gender, sexuality and drug use.

Methods

Study design and setting

This cross‐sectional study surveyed a convenience sample of attendees at major musical festivals in NSW, Australia, between November 2019 and March 2020 (see reference [2] for more study details). Music festivals were selected on the basis of requiring an approved Safety Management Plan under the NSW 2019 Music Festivals Act, indicating that the festival was deemed at high risk of drug‐related harm by the Independent Liquor and Gaming Authority [17]. Permission to conduct surveys was obtained from festival organisers.

Measures

The survey was developed by investigators with experience in clinical toxicology, social sciences research, harm reduction, nursing and festival peer support. Survey questions included participant demographics, specifically age, gender and sexuality. Participants were asked to select their age range rather than exact age to maintain confidentiality. Participants were asked about whether they had used or planned to use illicit drugs that day. Participants were then asked to consider a hypothetical situation where, if they or a friend were unwell at that festival, what factors might stop them from seeking help. Potential barriers and their correlates were identified based on: (i) a review of the literature on barriers to help‐seeking and their correlates in the context of substance use disorder and drug overdose; (ii) anecdotal experiences from investigators involved in peer support; and (iii) information from a recent coronial inquest into deaths at festivals [4, 5, 6, 8, 13, 16]. These barriers were reviewed as part of the peer co‐design process to confirm face validity. Multi‐choice options included not knowing where to find help; not knowing how unwell someone might be; getting in trouble with police; friends or relatives finding out; or that no factors would stop them seeking help (no barriers). Participants could select more than one option, could select ‘rather not say’ and could enter ‘other’ barriers by free text.

Participants and procedures

Between six and 10 trained volunteers collected surveys at each festival. Volunteers were predominantly recruited from DanceWize NSW, a program of the NSW Users and AIDS Association that utilises a peer education model to provide on‐site harm reduction, festival rovers, supervised care interventions and drug education services. Volunteers were provided with standardised training and study information and were not volunteering for DanceWize NSW at that event.

A convenience sample of festival‐goers was approached by roving volunteers throughout the festival, excepting dance areas, to complete the survey. Due to age restrictions of the selected festivals, all participants were 18 years or older. The sole exclusion criterion was impairment due to intoxication as assessed by study personnel implementing the survey. Research Electronic Data Capture (REDCap) mobile app version 5.4.7 was used to deliver the survey via an electronic tablet which took approximately 10 minutes for the participant to self‐administer (participants were supported by researchers if clarification was required). The survey opened with an electronic consent form that explained the study and the de‐identified nature of responses. The survey was anonymous and ‘rather not say’ was an option for all questions. Participants were offered a lollipop for survey participation.

Approval for this study was granted by The University of Sydney Human Research Ethics Committee (HREC 2019/652).

Data analysis

Data cleaning steps were undertaken to identify participants who had conflicting responses. Some participants initially reported that they did not intend to use drugs, but then later reported an intention to use a specific drug. These cases were included in the drug consuming population. Some participants reported that ‘nothing would stop me seeking help’ but also ticked one of the four barriers. In these cases, we removed their positive response to ‘nothing would stop me seeking help’, resulting in exclusive categories between people reporting no barriers and people reporting one or more barriers to help‐seeking. Individuals who did not provide a response or responded ‘rather not say’ were excluded from regression analyses for the variable(s) in which their data were missing.

Following these data cleaning steps, descriptive statistics for categorical data were generated as percentages. Bivariable logistic regression was used to assess the association of each of the four barriers to help‐seeking and ‘nothing would stop me’ (i.e. no barriers) as outcome variables, separately. Explanatory variables comprised participant age (18–20, 21–25 or 26+ years), gender (male, female, diverse gender), sexuality (heterosexual or diverse sexuality) and drug use (yes, no). All individuals who reported having used drugs on the day of the festival or who intended to use drugs that day were included in the drug‐using population. Each barrier variable was independently treated as a binary outcome variable. Those who checked the box within the survey indicating they experienced a given barrier were compared with those who did not check the box for that barrier.

After performing bivariable analyses (odds ratio and 95% confidence interval [CI] reported), multivariable logistic regression was performed to assess the association of all explanatory variables with each of the four barriers to help‐seeking and ‘nothing would stop me’ as outcome variables separately. Variance inflation factor was measured and a value of less than five was considered an indication of non‐collinearity. The significance level was set at P < 0.05.

In order to understand the intersection between perceived barriers we formed UpSet plots using the UpSetR package to visualise the overlap and clustering of barriers [18].

Analyses were performed in IBM SPSS Statistics version 26 (Chicago, IL, USA) and R version 3.6.2 (2019‐12‐12).

Results

Festival characteristics

Festivals (n = 6) ranged in size from 6200 to 16,289 attendees (median of 11,256 attendees). There were minor variations in music genre across festivals, with the majority (n = 5, 83%) playing predominantly electronic dance music. All festivals were single‐day events, ranging from 6 to 12 hours in duration, commencing between 10:00 and 18:00, finishing between 22:00 and 01:00. Two festivals (33%) were held indoors and four (67%) outdoors. No drug checking services were available at any of the festivals sampled. As presented elsewhere [2], there was a median of 100 presentations to on‐site emergency medical services per 10,000 festivalgoers (range 30–200/10,000). Furthermore, a median of three per 10,000 (range 1–7/10,000) required transportation from the festival medical service to the hospital. No drug‐related deaths were recorded at any of the six festivals.

Sample characteristics

In total, 1229 participants completed the survey: 978 (79.6%) completed all questions, 251 (20.4%) had missing data fields. Four hundred and ninety‐two (40.0%) participants were aged 18–20 years, 533 (43.4%) 21–25 years and 196 (15.9%) 26 years or older, while eight (0.7%) did not provide their age. Male gender was identified by 611 (49.7%) participants, 574 (46.7%) identified as female, 31 (2.5%) as non‐binary/gender fluid, different identity or ‘rather not say’ (grouped as ‘diverse gender’ for our analysis) and 12 (1.0%) did not provide gender. Most (972; 79%) participants identified as heterosexual, 246 (20.0%) as lesbian, gay or homosexual, bisexual, queer, other identity or ‘rather not say’ (grouped as ‘diverse sexuality’ for our analysis) and 11 (1.0%) did not respond. 372 (30.3%) reported that they intended on using drugs that day or had already used illicit drugs that day prior to completing the survey, henceforth referred to as people using drugs.

Barriers to help‐seeking

At least one barrier to help‐seeking was reported by 83.2% (n = 1022). One hundred and twenty‐eight participants (10.4%) entered ‘rather not say’ in response to the barriers question and were excluded from analyses. Twenty‐five (2%) selected ‘other barriers’ and of these, eight (0.7%) entered a free‐text response. All free‐text responses were either nonsensical or not relevant (Table 1 and Figure 1).

Table 1

Counts and proportions

Barriers to seeking help	No. of total population (% a ), n = 1229	No. of those using drugs (% a ), n = 372
No barriers (nothing would stop me)	198 (16.1%)	57 (15.3%)
Yes barriers
Not knowing where to find help	253 (20.6%)	47 (12.6%)
Not knowing how unwell someone might be	212 (17.2%)	42 (11.3%)
Getting in trouble with police	402 (32.7%)	153 (41.1%)
Friends or relatives finding out	188 (15.3%)	70 (18.8%)
Rather not say	128 (10.4%)	51 (13.7%)
Reporting at least 1 barrier	940 (76.5%)	273 (73.4%)

Multiple answers could be selected.

Figure 1

UpSet plot—Barriers to help‐seeking

From UpSet analyses of people reporting more than one barrier to help‐seeking (Figure 1), the most common combination of barriers was fear of trouble with police and fear of friends and relatives finding out, a combination which was reported by 4% (n = 51) of participants. The next most frequent combination was not knowing how unwell someone was and not knowing where to find help (2%, n = 28) (Table 2).

Table 2

Bivariate and multivariate regression

	No. of barriers (nothing will stop me)		Not knowing where to find help		Not knowing how unwell someone might be		Getting in trouble with police		Friends or relatives finding out		At least one barrier
Independent variable	Unadjusted OR (95% CI)	Adjusted OR (95% CI)	Unadjusted OR (95% CI)	Adjusted OR (95% CI)	Unadjusted OR (95% CI)	Adjusted OR (95% CI)	Unadjusted OR (95% CI)	Adjusted OR (95% CI)	Unadjusted OR (95% CI)	Adjusted OR (95% CI)	Unadjusted OR (95% CI)	OR (95% CI)
Age, years
≥26	2.37 (1.55–3.61)***	2.43 (1.58–3.75)***	0.71 (0.46–1.10)	0.70 (0.45–1.11)	0.63 (0.39–1.01)	0.67 (0.41–1.09)	1.08 (0.75–1.54)	0.84 (0.56–1.22)	0.42 (0.24–0.73)**	0.38 (0.22–0.68)***	0.42 (0.28‐0.64)***	0.41 (0.27–0.63)***
21–25	1.36 (0.95–1.93)	1.42 (0.99–2.04)	0.99 (0.73–1.34)	1.02 (0.75–1.40)	0.82 (0.59–1.13)	0.23 (0.82–0.58)	1.00 (0.70–1.43)	0.87 (0.67–1.14)	0.78 (56–1.08)	0.74 (0.53–1.04)	0.74 (0.52–1.05)	0.70 (0.49–1.01)
18–20 (ref)	1	1	1.00	1.00	1.00	1.00	1.00	1.00	1.00	1.00	1.00	1.00
Gender
Male	0.58 (0.33–1.05)	0.72 (0.52–1.00)*	0.8 (0.43–1.50)	0.60 (0.45–0.82)***	0.63 (0.33–1.20)	0.78 (0.57–1.07)	1.41 (0.90–2.22)	1.31 (1.01–1.70)*	2.13 (1.16–3.91)*	1.52 (1.09–2.14)*	1.33 (0.97‐1.81)	1.38 (1.00–1.91)*
Diverse	0.79 (0.16–3.85)	0.37 (0.08–1.63)	Not enough power	Not enough power	Not enough power	Not enough Power	1.73 (0.50–5.97)	0.83 (0.35–1.98)	5.39 (1.48–19.63)*	3.07 (1.22–7.68)*	2.12 (0.63‐7.14)	2.71 (0.61–11.92)
Female (ref)	1	1	1.00	1.00	1.00	1.00	1.00	1.00	1.00	1.00	1.00	1.00
Sexuality
Diverse	0.99 (0.67–1.44)	0.93 (0.62–1.39)	0.96 (0.68–1.37)	0.90 (0.62–1.31)	0.57 (0.37–0.87)**	0.55 (0.35–0.86)**	1.26 (0.93–1.70)	1.40 (1.02–1.93)*	0.78 (0.51–1.17)	0.77 (0.49–1.21)	1.01 (0.69–1.49)	1.08 (0.72–1.62)
Heterosexual (ref)	1	1	1.00	1.00	1.00	1.00	1.00	1.00	1.00	1.00	1.00	1.00
Drug use
Yes	0.99 (0.70–1.39)	0.94 (0.66–1.35)	0.49 (0.34–0.69)***	0.49 (0.34–0.71)***	0.55 (0.38–0.79)***	0.56 (0.38–0.82)**	1.94 (1.49–2.52)***	1.92 (1.47–2.52)***	1.57 (1.13–2.19)**	1.48 (1.05–2.09)*	1.01 (0.72‐1.42)	1.06 (0.74–1.5)
No (ref)	1	1	1.00	1.00	1.00	1.00	1.00	1.00	1.00	1.00	1.00	1.00
Alcohol consumption
Yes	0.86 (‐0.60–1.22)	0.82 (0.57–1.18)	1.18 (0.84–1.67)	1.14 (0.81–1.62)	0.72 (0.51–1.02)	0.70 (0.5–0.99)*	1.05 (0.78–1.40)	1.10 (0.82–1.49)	0.86 (0.60–1.24)	0.93 (0.64–1.34)	1.17 (0.82–1.67)	1.22 (0.85–1.75)
No (ref)	1	1	1	1	1	1	1	1	1	1	1	1

P ≤0.05.

P ≤0.01.

P ≤0.001.

One hundred and ninety‐eight participants (16.1%) reported that nothing would stop them seeking help. In multivariate analysis, people ≥26 years old were more likely to report that nothing would stop them help‐seeking compared to people 18–20 years old (adjusted odds ratio [AOR] 2.43, 95% CI 1.58–3.75).

The most commonly reported barrier was fear of getting into trouble with the police, reported by 32.7% (n = 402) of participants. In multivariable analyses, those grouped as diverse sexuality (AOR 1.40, 95% CI 1.02–1.93) and people using drugs (AOR 1.92, 95% CI 1.47–2.52) were more likely to report fear of getting into trouble with police as a barrier to help‐seeking than their referent populations (heterosexual people and people not using drugs, respectively).

Not knowing where to find help was the second most frequently reported barrier, reported by 20.6% (n = 253) of participants. People reporting drug use (AOR 0.49, 95% CI 0.34–0.71) were less likely to select this barrier compared to those not reporting drug use.

Seventeen percent (n = 212) of participants reported not knowing how unwell someone might be as a barrier to help‐seeking. In multivariate analysis, those grouped as diverse sexuality (AOR 0.55, 95% CI 0.35–0.86) and those reporting drug use (AOR 0.56, 95% CI 0.38–0.82) were less likely to report not knowing how unwell someone might be as a barrier compared with their reference populations (heterosexual people and those not reporting drug use, respectively).

Concern about other friends or relatives finding out was the least frequently reported barrier (15.3% [n = 188] participants). In multivariate analysis, people 26+ years old (AOR 0.38, 95% CI 0.22–0.68) were less likely to report this concern than people 18–20 years old. Males (AOR 1.52, 95% CI 1.09–2.14) and people of diverse gender (AOR 3.07, 95% CI 1.22–7.68) and those using drugs (AOR 1.48, 95% CI 1.05–2.09) were more likely to report this concern, compared with their referent populations (females and those not using drugs, respectively).

Discussion

This is the largest contemporary in situ survey to be completed on‐site over one festival season in Australia and, to our knowledge, only the second to report specifically on barriers to help‐seeking among festival attendees [16]. The majority (over 80%) of participants reported at least one barrier to help‐seeking in the hypothetical event of themselves or a friend becoming unwell at a festival. Fear of getting into trouble with police, and not knowing where or when to seek help, were the most commonly reported barriers.

Fear of getting in trouble with the police is a barrier consistently reported across studies examining barriers to seeking medical attention in the context of an overdose among people who use drugs [5, 8], including in festival environments [16]. In our study, this barrier was more likely among people taking drugs and those who indicated diverse sexualities. Consistent with this, previous studies in the US and Australia have found that the LGBTQ+ community continues to have disproportionately negative experiences with police with the potential to negatively impact on help‐seeking for drug‐related problems [19, 20]. While there are notable examples of initiatives aimed at improving this relationship [21, 22], our findings suggest a more concerted effort is required.

Help‐seeking has been correlated with policing strategies in a recent study across 20 countries; individuals from countries with prohibition‐based drug policies reported a greater propensity for fear of criminal sanctions being a barrier to help‐seeking for drug or alcohol treatment [7]. Almost half of an Australian sample of people who use heroin reported that there were factors that had delayed or stopped them seeking medical assistance, the most common impediment being a fear of police involvement [5]. There are examples of policy to re‐dress this issue, such as the Canadian Good Samaritan Drug Overdose Act of 2017, providing immunity from prosecution for simple drug possession to overdose victims or bystanders who phone for emergency assistance. However, a subsequent qualitative study found that, despite this legislation, cumulative negative experiences with police continued to influence as an ongoing barrier [23]. The recent Coronial Inquest into Deaths at Festivals in NSW, Australia [13] identified prohibitionist policing strategies as one of the main factors associated with delayed help‐seeking and subsequent mortality and recommended rethinking this approach. While on‐the‐spot fines were introduced in early 2019 in NSW as an alternative to criminal prosecution for people in possession of a small quantity of prohibited substances at music festivals [24], experience in other countries and an ongoing prohibitionist approach suggest that this will be insufficient to re‐dress this barrier to help‐seeking among people who use drugs. Further research into the impact of different policing strategies and cultures on these barriers, both in different jurisdictions across Australia and overseas, may better inform policing policy and decision‐making to promote help‐seeking behaviours and reduce the risk of harm.

Not knowing where to find help and not knowing how unwell someone was were common barriers to help‐seeking among our sample. Both were associated with reporting abstinence from drugs at the festival, whereas those of diverse sexualities appeared to be more confident at identifying how unwell someone was. It is possible that participants reporting to be abstaining from drug use in our study have limited experience with drug taking themselves or are part of social groups in which drug taking is not prevalent. Past studies have demonstrated that a large amount of knowledge relating to drug overdose is acquired through experience [25]. It may therefore be reassuring that people who take drugs are more likely to report knowing where to find help and how unwell someone is. Our findings are consistent with a 2019 retrospective Australian survey of people who regularly use ecstasy and other illicit stimulants which found that, of those who reported attending festivals, 87% were aware of on‐site medical services at the last festival they attended [26]. However, previous studies have also demonstrated that people not using drugs are less likely to report barriers to help‐seeking following drug overdose [4] and so there is value in addressing this knowledge gap. Live, on‐site public health messages, peer dissemination of knowledge, pre‐event messaging, social media posts, festival signage, performer announcements and strategic positioning and signposting of emergency medical services during a festival may be helpful. Other innovative measures such as objective wellness resources, for example, wearable temperature monitoring devices, should be explored.

Evidence also suggests that knowledge relating to recognising and responding to drug use is retained within social groups; our study suggests this may be particularly the case for the LGBTQ+ community. However, further work is required to determine the accuracy of these perceptions. Past studies have identified that while endowing knowledge, past experience with drug overdose may also serve as a barrier to help‐seeking. Those involved may attempt to ‘ride it out’ or attempt supportive measures themselves prior to seeking medical attention, including administering other recreational drugs as counter measures [6, 8]. Enhanced peer messaging around the importance of help‐seeking, as well as addressing the other barriers identified here may help to address this issue.

The youngest participants, men, people of diverse genders and those using drugs were most likely to list fear of friends and relatives finding out as a barrier. This may reflect a combination of concerns about confidentiality (e.g. that health workers may disclose information to individuals' friends or relatives or to police), stigma and other sociocultural issues. The recent Coronial Inquest into Deaths at Festivals underscored the importance of appropriate school‐based education programs to de‐stigmatise drug use [13]. There are also initiatives aimed at facilitating open family dialogue about drug use [27], however, further work exploring the foundations of these concerns is warranted.

Limitations

Limitations of this research include the use of a hypothetical situation to assess help‐seeking in the event of someone becoming unwell but not specifying the cause of their becoming unwell (i.e. not specifically drug‐related illness). Hypothetical choices may not align with actual behaviour [28]. Information about prior experiences of needing help in festival environments and barriers preventing this would have been valuable, but were not interrogated and were beyond the scope of this study. Data collected were self‐report only: self‐report data are subject to recall problems and social desirability bias. Only NSW festivals requiring a festival licence were included. Findings may differ at other NSW festivals, and in other jurisdictions where policing strategy and culture may be different.

Conclusions

Information gathered here has permitted the identification of barriers to help‐seeking, and the demographic groups at increased risk of delayed engagement with health services; such delays are likely risk factors in themselves for morbidity and mortality. Building upon this research, it is vital that we deepen our understanding of barriers to help‐seeking among populations at risk of drug‐related harms across varied settings. This will permit optimisation of health and policing policy and facilitate the direction of resources to reduce the risk of harms. In addition to expanding the knowledge base around these barriers, this study further emphasises the potential consequences of law enforcement strategies at festivals that have been linked to hazardous drug use behaviour [13, 14, 15] in Australia. Policing and drug policy strategies should be examined to establish whether alternative methods may reduce these negative impacts. Further investigation into the drivers of these barriers, alongside work to minimise or eliminate them, should be part of ongoing work to reduce drug‐related harms in festival environments.

Conflict of Interest

AP has received a untied educational grant from Seqirus and Mundipharma for the study of opioid medications. AP is supported by an NHMRC Investigator Fellowship (#1174630). JB is supported by an NHMRC Investigator Grant (#1196560). All other researchers had no interests to declare.

Table A1

Counts: Help‐seeking behaviours

Barrier	No barriers (nothing would stop me)		Not knowing where to find help		Not knowing how unwell someone might be		Getting in trouble with police		Friends or relatives finding out
	No drug use	Drug use	No drug use	Drug use	No drug use	Drug use	No drug use	Drug use	No drug use	Drug use
Gender
Male	60	27	75	26	72	21	120	94	62	47
Female	79	28	124	21	92	21	124	52	53	17
Diverse	1	2	7	0	5	0	3	6	3	5
Age, years
≥26	35	14	25	7	19	7	30	33	5	12
21–25	62	24	89	24	65	23	103	67	45	34
18–20	44	19	92	16	86	12	116	53	68	24
Sexuality
Diverse	23	23	43	7	22	7	65	27	17	15
Heterosexual	118	40	162	39	148	34	184	126	101	54
Police presence influence decision to take drugs
Yes	NA	9	23	11	24	10	59	48	32	22
No	NA	48	183	36	146	32	190	105	86	48
Drug used
MDMA	NA	48	NA	34	NA	30	NA	117	NA	58
Non‐MDMA drug use	NA	9	NA	13	NA	12	NA	36	NA	36
Pre‐loading
Yes	NA	25	NA	12	NA	12	NA	45	NA	15
No	NA	32	NA	35	NA	30	NA	108	NA	55
First time use
Yes	NA	2	NA	4	NA	5	NA	4	NA	8
No	NA	55	NA	43	NA	37	NA	145	NA	64