Navigating social norms of injection initiation assistance during an overdose crisis: a qualitative study of the perspectives of people who inject drugs (PWID) in Vancouver, Canada

Abstract

Despite the proliferation of fentanyl and fentanyl-adulterated opioids in North America, the impacts of this drug market change on injection initiation processes have not been examined. With the aim of informing structural interventions to address injection initiation and related harms, we explore how people who inject drugs (PWID) in Vancouver, Canada understand and navigate social norms of initiating others into injecting within the context of an overdose crisis. In-depth qualitative interviews were conducted with 19 PWID who reported helping someone inject for the first time. Participants were recruited from two cohort studies of PWID. Participants articulated moral dilemmas about assisting others with injecting. While participants described a ‘moral code’ prohibiting assisting injection-naïve individuals, this code was not the sole consideration shaping social action around injection initiation. Rather, PWID exercised agency about whether and how to assist novice injectors within the context of constraining and enabling social norms around practicing interpersonal responsibility. Changes to the drug market heightened feelings of moral culpability and criminal liability among PWID who assisted others into injection, given that injecting heightened initiates’ risk of overdose. These concerns operated in tension with the aim of protecting novice injectors from harms associated with an increasingly potent and unpredictable drug supply by providing them with injection assistance, education and supervision. Our analysis of how PWID practice interpersonal responsibility helps conceptualise how ‘moral codes’ prohibiting initiation assistance are managed and negotiated amidst structural vulnerability. Structural interventions reducing the vulnerability of novice injectors should be prioritized, including the implementation of supervised injection sites allowing for assisted injection, Good Samaritan laws, and policy changes conducive to a safer drug supply.

Introduction

Illicit drug overdose has rapidly become a leading cause of accidental death in the US and Canada, a trend driven primarily by the proliferation of highly potent fentanyl and fentanyl-adulterated substances within the illicit drug market (Centres for Disease Control and Prevention, 2018; Statistics Canada, 2018). Fentanyl adulteration has been accompanied by localized overdose epidemics in regions located along the distribution network for fentanyl, including hotspots in mid-Atlantic and Midwestern United States, and Western Canada (Centers for Disease Control and Prevention, 2018; Public Health Agency of Canada, 2018). In British Columbia, Canada, illicit drug overdoses accounted for an estimated 1450 deaths in 2017, which represents a 7-fold increase from overdoses recorded in 2010 (British Columbia Coroners Service, 2018). Among those who died from illicit overdose in BC between 2016 and 2017, fentanyl was the most frequently detected substance (76% of deaths) and injection was the most common mode of consumption (41%) (British Columbia Coroners Service, 2018).

Preventing overdoses and other harms associated with injection drug use is an urgent public health priority. Efforts to reduce injection-related morbidity and mortality include interventions to promote non-injecting routes of administration (e.g. smoking) and prevent transitions into injection drug use (Bridge, 2010; Dolan et al., 2004), a route of administration which is associated with overdose and other harms (e.g. HIV, HCV, skin and soft tissue infections) (Degenhardt et al., 2017). As such, public health experts and international organizations such as UNAIDS and the United Nations Office on Drugs and Crime have called for allocation of resources toward preventing injection initiation (United Nations Office on Drugs and Crime (UNODC), 2004; Vlahov, Fuller, Ompad, Galea, & Des Jarlais, 2004).

Previous research indicates that people who inject drugs (PWID) commonly require and receive assistance from established injectors during their first injection due to the complexity of preparing and injecting illicit substances (Crofts, Louie, Rosenthal, & Jolley, 1996; Goldsamt, Harocopos, Kobrak, Jost, & Clatts, 2010; Harocopos, Goldsamt, Kobrak, Jost, & Clatts, 2009; Rotondi et al., 2014; W. Small, Fast, Krusi, Wood, & Kerr, 2009). Although prevalence studies find that less than half of PWID report ever facilitating injection initiation (Bryant and Treloar, 2008; Crofts, et al., 1996), those who do assist have reported assisting multiple people with their first injection (Rotondi, et al., 2014). Given these findings, efforts to prevent initiation of injection drug use largely entail behavioural interventions focused on dissuading PWID from initiating others into injecting through educational outreach with established injectors (Hunt, Stillwell, Taylor, & Griffiths, 1998; Strike et al., 2014). Evaluations of such injection initiation prevention programs in the United Kingdom (Break the Cycle) and Canada (Change the Cycle) suggest short-term reductions in the number of initiation events reported by participants (Hunt, et al., 1998; Strike, et al., 2014), however the peer-reviewed evidence supporting these interventions is limited (Werb et al., 2013).

Sociological and anthropological literature on injection initiation suggests that the process involves negotiations of moral boundaries and social norms (Guise et al., 2018; Rhodes et al., 2011; Small, et al., 2009; Wenger, Lopez, Kral, & Bluthenthal, 2016). Transitions into injection drug use have been narrated by some PWID as a crossing of a moral boundary and transition into a new symbolic identity (Fitzgerald, Louie, Rouen, & Crofts, 1999; Martin, 2010; Rhodes, et al., 2011). Qualitative research from various settings indicates that while PWID commonly assist others in their drug using networks, the initiation of novice injectors is perceived to be a moral boundary that should not be transgressed (Guise, et al., 2018; Rhodes, et al., 2011; Small, et al., 2009; Wenger, et al., 2016). Small and colleagues (2009) note how this boundary is upheld in Vancouver’s Downtown Eastside neighbourhood by a ‘moral code’ among PWID which stipulates that young people who use drugs should not be initiated into injection drug use. This moral code might prevent or delay transitions to injection drug use if it effectively deters established injectors from initiating novice injectors.

In practice, however, PWID routinely transgress this moral code to meet survival needs (e.g., offering assistance in exchange for drugs, sex, money or other material goods), fulfil social obligations (e.g., to help others achieve a better and more ‘efficient’ high) and reduce harms incurred by inexperienced injectors (e.g., overdose, missed injections, infections) (Fairbairn, Small, Van Borek, Wood, & Kerr, 2010; Kolla et al., 2015; Small, et al., 2009; Wenger, et al., 2016). As such, there is a growing recognition that individual-level injection initiation prevention programs must be complemented with interventions that address structural and environmental conditions that contribute to injection initiation processes (Werb et al., 2018). Specifically, understandings of injection initiation practices must be situated with consideration for the structural vulnerability of PWID, particularly low-income and racial/ethnic-minority PWID, who are subject to mutually-enforcing economic, political and cultural forces that relegate them to the bottom of social hierarchies, predispose them to poor health, and constrain their agency within processes of injection initiation (Guise, Horyniak, Melo, McNeil, & Werb, 2017; T. Rhodes et al., 2012).

To that end, drug markets themselves are structural forces which may powerfully shape social norms and practices around injection initiation. Public health researchers increasingly recognize the importance of drug market characteristics (e.g. types of drugs available, composition, purity and price) as structural factors that pattern injection drug user practices and harms (Andrade and Sifanech, 1999; Ciccarone, 2017; Ciccarone, Ondocsin, & Mars, 2017; Mayer et al., 2018; Strang, P., & Gossop, 1997; Swift, Maher, & Sunjic, 1999; van Ameijden and Coutinho, 2001; Zafar and Hasan, 2002). Ciccarone (2009), for example, suggests that the dominance of Mexican-sourced black tar heroin in Western US states may encourage transitions into injecting, as the substance is difficult to inhale given its composition (Ciccarone, 2009). In contrast, the emergence of high purity brown Asian heroin in the Netherlands has been linked to uptake of ‘chasing the dragon’ (heroin smoking) as a viable alternative to injection(Grund and Blaken, 1993). Drug market shifts may also precipitate transformations to social norms around injection initiation, as seminal ethnographies of street-based drug scenes demonstrate how moral expectations about appropriate behavior are embedded within broader moral economies of drug markets(Bourgois, 1998; Karandinos, Hart, Castrillo, & Bourgois, 2014; Wakeman, 2016).

The emergence of fentanyl within North American drug markets represents a significant transformation within the risk environment of PWID, with potential implications for how PWID understand the risks and morality of assisting novice injectors. Highly potent fentanyl and fentanyl-related analogues have rapidly flooded the heroin supply in the US and Canada (Ciccarone, 2017; Drug Enforcement Administration Diversion Control Division, 2016). In Vancouver, a 2018 drug-checking pilot project found that fentanyl was present in 90.6% of all opioid samples submitted for testing at two downtown supervised injection sites (Tupper, McCrae, Garber, Lysyshyn, & Wood, 2018). The dramatic rise of fentanyl and fentanyl-related overdoses is important to consider in relation to injection transitions, as the increased risk of overdose and uncertainties surrounding drug potency contribute new ethical tensions concerning injection initiation assistance that may reinforce or disrupt existing social norms around the practice.

Despite the rapid proliferation of fentanyl and widespread adulteration in North America, the impacts of these drug market changes on the social norms of injection initiation have not to our knowledge been examined. A better understanding of the dynamic between drug supply and injection initiation norms will help inform development of context-specific structural interventions to prevent injection initiation and reduce harms associated with injection drug use, including transmission of HIV and HCV. In this study, we examine how PWID in Vancouver, British Columbia understand and navigate the social norms of injection initiation within the context of an overdose crisis driven by fentanyl-adulteration, with particular attention to how PWID narrate moral codes prohibiting initiation assistance. This question has not yet been studied and has pressing public health importance in North America and other regions where fentanyl-adulteration is driving overdose-related morbidity and mortality.

Methods

This study was undertaken as part of a multi-site ethno-epidemiological study investigating the potential of socio-structural interventions (e.g., opioid agonist therapy, supervised injection facilities, stable housing) to disrupt processes of injection initiation assistance. Ethno-epidemiology draws on epidemiological and qualitative methods to uncover how social and structural processes pattern health and drug harms, with the aim of informing development of interventions that minimize harm (Lopez et al., 2013; Wagner et al., 2012). Vancouver, Canada is one of seven sites included in the study, along with San Diego, USA; Tijuana, Mexico; and Paris, Marseille, Bourdeux, and Strasbourg, France (Werb et al., 2016). In this paper, we report on findings from the qualitative component of this study exploring people’s experiences with injection initiation assistance in Vancouver, Canada, the only study site with widespread fentanyl adulteration. Between January and April 2017, we conducted 19 semi-structured interviews with people who reported ever helping someone inject for the first time. All study activities were approved by the Research Ethics Boards of the University of British Columbia and Providence Health Care.

A purposive sampling strategy was used to identify PWID who had experience assisting others with injecting for the first time. Participants were recruited through the AIDS Care Cohort to Evaluate Exposure to Survival Services (ACCESS) and the Vancouver Drug User Study (V-DUS), two prospective cohort studies based in Vancouver composed of HIV-positive (ACCESS) and HIV-negative (V-DUS) PWID (Strathdee et al., 1998; Wood et al., 2009). Participants were selected for recruitment through these cohorts if, during follow-up interviews conducted after January 2016, they reported ever having helped someone inject who had never injected before. Eligible participants were invited over the phone to participate in the study and consent was obtained in-person prior to the interviews. Participants were provided a $30 honorarium (CAD) for their participation. In-depth interviews were conducted at a storefront research office located in Vancouver’s Downtown Eastside neighborhood and facilitated using an interview guide informed by past qualitative work on injection initiation (Guise, et al., 2017) and fieldwork recently completed in San Diego and Tijuana (Guise, et al., 2018). This interview guide was structured to explore participants’ injection drug use histories, including their own experiences of first injecting drugs and providing assistance to injection-naïve individuals. Interviews were audio-recorded and transcribed verbatim for analysis. Recruitment continued until saturation was reached in data, whereby the research team determined that no new themes or information were emerging from additional interviews (Grady, 1998).

After data collection had ended, the lead author reviewed all interviews and drafted analytic notes. Her first review and coding of interview transcripts highlighted fentanyl-adulteration in the drug supply as a socio-structural factor which was framing participants’ narratives of injection initiation assistance. An initial coding framework was then constructed by MO and RM, organized around social meanings ascribed to initiation assistance within the context of fentanyl proliferation and the overdose crisis. We imported interview transcripts into NVivo (version 11) to facilitate data management and coding. To maintain anonymity, all participants were assigned pseudonyms using an online pseudonym generator.

Interviews were analyzed using an abductive approach in which data were reviewed and coded using existing theory, with the aim of reconstructing theory based on new insights from the data (Buroway, 1991). Following Goffman’s dramaturgical theory, we view the interview encounter as an interaction in which the interviewee is performing a narrative of self that highlights aspects of their identity most likely to garner sympathy, administration and esteem, while concealing those aspects that could be interpreted as defects or weaknesses (Goffman, 1973).This approach directs us to interpret the interview itself as a performance in which a member of a stigmatized group seeks to articulate a presentation of self that is socially acceptable to an interviewer situated within a public health research institution and within the broader context of an unfolding public health emergency. This approach does not diminish the ‘truth’ of the interviewee’s account, but rather encourages consideration for how participants narrate a responsible self within this particular interaction.

We drew on the ‘risk environment framework’ and concepts of ‘moral codes’ and ‘structural vulnerability’ to interpret themes about injection initiation emerging from analysis. The ‘risk environment’ describes the overlapping social, physical, economic and policy environments in which forces exogenous to the individual interact to socially pattern harms (Rhodes, 2002). The complementary concept of ‘structural vulnerability’ refers to the vulnerability of individuals to disease and poor health produced by their subordinated status in social hierarchies (Quesada, Hart, & Bourgois, 2012). These concepts focus analyses on how an individual’s risk environment, operating at the micro, meso- and macro-level, structures the agency of PWID and their vulnerability to drug-related harms (Rhodes, 2009; Rhodes, et al., 2012). In our analysis, we apply these concepts to examine fentanyl-adulteration and proliferation as a feature of the risk environment that shapes vulnerability to drug-related harm and social norms around injection initiation. We are particularly interested in how macro-level drug market changes interact with ‘moral codes’—a set of injunctive social norms which operate at micro and meso levels to proscribe how people ought to behave (Jimerson and Oware, 2006)—to constrain and enable the agency of PWID in injection initiation processes. We distinguish moral codes from other social norms governing injection practices—for example, norms around drug preparation and the order of injecting—by the moral significance commonly attributed such practices (i.e. as being ‘morally wrong’ or ‘bad’), and associated feelings that might accompany transgressions of these codes (Guise, et al., 2018).

Findings

Table 1 provides a breakdown of demographic information for all 19 interview participants, which included 9 cis-gender women, 8 cis-gender men and 1 transgender man. Our interviews revealed a discordance between the injection assistance practices reported in longitudinal cohort surveys and those described within the qualitative interviews. While all participants had previously reported ever helping someone inject for the first time during cohort interviews, nearly half (n=7) explicitly denied providing initiation assistance when asked about these experiences during the qualitative interviews (see Table 1). In the analysis that follows, we explore how this discordance may be linked to social norms governing injection initiation which are particularly salient within the context of the fentanyl-driven overdose crisis. We further explore how fentanyl adulteration frames how social norms around injection initiation are narrated by PWID in Vancouver’s Downtown Eastside.

Table 1.

Primer qualitative interview participants in Vancouver, BC (n=19)

ID	Pseudonym	Gender	Age	Ethnicity	Substances used in past 30 days (drug of choice in bold)	Route of administration to consume drugs in past 30 days	Reported ever providing initiation assistance in interview	Housing status reported in interview	Self-reported health issues
1	Paul	M	62	Indigenous	Heroin, Cocaine, Opiates, Alcohol	Inject Ingest	Denies assisting	Public SRO	HCV
2	Susan	F	57	White	Cocaine, Marijuana, Prescribed morphine	Inject	Denies assisting	House/Apartment	HCV, Chronic pain
3	Florence	F	45	Indigenous	Crack cocaine, Marijuana	Inject, Inhale	Denies assisting	Public SRO	TB
4	Arthur	M	34	Indigenous	Crystal Meth, Marijuana, Alcohol	Inject, Inhale, Ingest	Assisted an acquaintance/ stranger	Private SRO	HIV, HCV
5	Sarah	F	32	Indigenous	Crystal meth, cocaine	Inject, Inhale, Snort	Assisted others, but unsure if first time	Public SRO	HCV
6	Tammy	F	35	Indigenous	Crystal meth, Marijuana, Alcohol	Inject, Inhale, Ingest	Assisted her boyfriend	Public SRO	HCV
7	Sadie	F	48	Indigenous	Crack cocaine, Marijuana, T3 prescription	Inhale	Denies assisting	Public SRO	Bipolar
8	Jason	M	36	White	Crystal Meth, Heroin, Opiates, Marijuana, Alcohol	Inject, Inhale	Denies assisting	Public SRO	HCV
9	James	M	37	Indigenous	Crystal Meth, Marijuana, Alcohol	Inject, Inhale, Snort, Ingest	Assisted other, but unsure if first time	House/Apartment	None
10	Barbara	F	36	Indigenous/Metis	Heroin, Crystal Meth, Marijuana, Crack Cocaine, Alcohol	Inject	Assisted brother	Private SRO	HIV, HCV
11	Kevin	M	34	Chinese and Filipino	Heroin, Cocaine, Marijuana, Opiates	Inhale	Assisted 3–4 times	House/Apartment	None
12	Donna	F	57	White	Heroin, Cocaine, Crack Cocaine, Opiates, Crystal Meth	Inject, Inhale	Assisted a friend	Public SRO	Hip replacement
13	Thomas	TM	31	White	Crystal Meth	Inject	Assisted brother	House/Apartment	HIV
14	Angela	F	47	White	Cocaine, Heroin, Crack Cocaine, Opiates	Inject, Inhale	Assisted an acquaintance/stranger	Private SRO	HBV, HCV
15	Vernon	M	43	Indigenous/Metis	Marijuana, Crack Cocaine, Crystal Meth, Atavan	Inject, Inhale	Denies assisting	Public SRO	Abscesses, HBV, HIV, Arthritis
16	Doug	M	46	Indigenous	Cocaine, Heroin, Crystal Meth	Inject	Denies assisting	Public SRO	Abscesses, HIV, HCV, Hep B, TB
17	Matthew	M	54	White	Heroin, Marijuana, Crack Cocaine, Opiates, Benzos	Inject, Inhale	Assisted a friend	Public SRO
18	Mary	F	41	Indigenous, Black	Heroin, Crystal Meth, Alcohol	Inject	Assisted intimate partner, friend	House/Apartment	HCV, Psoriasis
19	Alexandra	F	35	Indigenous	Heroin, Crystal Meth	Inject	Assisted an acquaintance	Sober living facility	HCV

M=Male, F=Female, T=Trans; SRO=Single Room Occupancy Hotel; HBC: Hepatitis B Virus; HCV: Hepatitis C Virus; HIV: Human Immunodeficiency Virus; TB: Tuberculosis

Fentanyl-adulteration and the moral code against injection initiation

Participants’ accounts emphasized social norms against assisting injection-naïve individuals. Among both those who reported and denied assisting others for the first time during qualitative interviews, initiation assistance was positioned as morally wrong. The prohibition against initiation assistance was often described as a moral code that PWID had to learn and abide to: “I found out later it’s a rule that you don’t fix somebody for the first time.” (‘Donna,’ white woman, age 57). Reflecting on her own experience assisting her boyfriend for the first time, ‘Sadie,’ a 48-year-old Indigenous woman, recalled: “It was just something that was taboo that you didn’t do. To fix somebody else.” This moral code was most salient with respect to youth, as there was a common perception that providing initiation assistance to youth was ethically unacceptable. As ‘Kevin,’ a 34-year-old Chinese and Filipino man, stated: “I don’t condone them teaching kids how to inject themselves because that’s just not appropriate.”

For some participants, the proliferation of fentanyl in the drug supply reinforced the importance of the code against assisting others for the first time. The uncertainty around drug potency meant participants were reluctant to assist strangers with injecting heroin or other drugs potentially contaminated with fentanyl, even if these strangers claimed to have previous injection experience, given the high risk of fatal overdose associated with fentanyl injection. Within the context of a fentanyl-contaminated opioid supply, injection initiation assistance was sometimes described as playing a game of “Russian Roulette.” In reflecting on their experiences of assisting with injection initiation, some participants identified fentanyl-adulteration as a primary reason for a reduction in their willingness to assist others for the first time. ‘Vernon,’ a 43-year-old Indigenous man, reflects on how his feelings toward assisting injection naïve individuals changed along with fentanyl-proliferation:

I wouldn’t really want to enable them or take that risk, especially nowadays with the fentanyl. Maybe back in ‘92 it might be different, because there wasn’t all this fentanyl and carfentanil and stuff. But with the risk factor now, no, I wouldn’t do it.

As James, a 37-year old Indigenous man, explained: “I won’t inject someone with heroin because I don’t want nothing to do with Fentanyl, man.”

Practicing and performing interpersonal responsibility

While participants described injection initiation assistance as a prohibited practice, detailed description of participants’ own involvement in and witnessing of injection initiation suggested that the permissibility of these practices was not determined solely by the moral code they described. PWID also exercised agency about when and how to assist others. These decisions were negotiated in relation to norms around practicing and performing interpersonal responsibility. That is, participants articulated a framework of interpersonal responsibility that could be defined as an interpersonal duty to help others use drugs ‘responsibly.’ Interpersonal responsibility also implies that responsibility for adverse outcomes (e.g. overdose) is collectively shared and managed by those involved in the injection process. This framework of interpersonal responsibility was core to ethical practice around injection assistance. Whereas participants’ emphasized their own agency in deciding to inject, they tended to emphasize the responsibility of those who assist when describing others’ transitions into injection use. Doug, a 44-year-old Indigenous man, rejected the notion that new injectors were individually responsible for their decision to inject, and suggested a framework of interpersonal responsibility for injection initiation in which those who assist are also responsible for new injector’s subsequent drug-related harms:

Doug: I know a couple of people that have done that [assisted novice injectors], and I tell them the same thing. ‘Why did you do that, man? Like fuck, you just wrecked that kid’s life, you know.’

Interviewer: Yeah. What do they say?

Doug: [They say] ‘What are you talking about? I didn’t wreck his life. He wrecked his own life. He’s down here doing dope.’ [I say,] ‘You stuck the needle in his arm for the first time. You stuck it in his arm. And then now he’s going to want you to stick it in his arm all the time.’

Participants sought to perform interpersonal responsibility in the interview by strategically minimizing their role in injection initiation or highlighting other responsible practices they engaged in during the injection initiation process. We use the term ‘perform’ here to capture how individuals narrate their responsibilities to other actors. The most prominent technique employed by participants to assert their interpersonal responsibility within the interviews was to emphasize how, even if they went on to assist novice injectors, they would first actively try to dissuade new injectors who requested their help. Questions about their experience assisting novice injectors elicited detailed and dramatized stories of participants’ efforts to dissuade new injectors, which were often accomplished by cautioning novice injectors of the stigmatized mantle of being a ‘junkie’ they would assume if they started injecting. In some cases, participants described emphasizing track marks and abscesses on their own bodies as an example of the embodied forms of suffering that new injectors should avoid. Doug recounts the following tactic to dissuade a novice injector:

Doug: I told her, I says, ‘Look, you see that?’

Interviewer: Yeah. You’re showing her your track marks.

Doug: At the time, my veins were like little train tracks all the way up and all the way down.

Interviewer: Were they like infected or something?

Doug: And I said, ‘Look at this. Do you want this on you every day, you know? Do you want to look at that? You’ll never be able to wear a short-sleeved shirt […] don’t use, because like you don’t want to end up like 90 percent of these people down here that don’t give a shit about themselves. You’ll be toothless and hairless in a week. You’ll be full of scabs and pimples and whatever.’

Another way in which participants performed interpersonal responsibility in the interview encounter was by emphasizing the challenges of distinguishing novice injectors from experienced injectors. While some participants explicitly denied having assisted others for the first time, others provided a more ambiguous response, asserting that they may have assisted someone for the first time, but did not or would not know if it was someone’s first time. Notably, all participants denying assistance in this way were women, suggesting a gendered difference in how responsible practice is understood and narrated. These stories align with participants’ own experiences of injecting for the first time, as some described obtaining help by purposefully not disclosing that it was their first time.

There were also a few cases in which initiation assistance was described as occurring under coercive circumstances. These circumstances prevented participants from behaving in ways they considered responsible. For some participants, particularly amongst women, this coercion occurred within the context of intimate and close relationships in which threats of violence or other harm compelled them to provide injection initiation assistance. Tammy, a 35-year-old Indigenous woman, describes feeling like she had “no choice” when her boyfriend asked for assistance fixing for the first time, as he had previously become “violent and pissed off when he doesn’t get his way.” Alexandra, also a 35-year-old Indigenous woman, similarly recounted feeling “forced into” assisting a friend’s girlfriend after that woman “cornered me in her SRO [single room occupancy housing] and screamed at me for like 20 minutes that I had to do it, and I wasn’t leaving the room until I did it.” James also described feeling like he had “no choice” but to assist others after being pressured “to the point where I’m gonna get hurt real bad if I don’t.”

These examples together demonstrate the significance of practicing interpersonal responsibility within participants’ narratives about injection initiation. Further, they suggest that social norms around interpersonal responsibility frame how PWID manage and negotiate moral codes prohibiting injection initiation, including how they enact moral codes in practice and how they make sense of transgressions of these codes.

Fentanyl-adulteration and perceived legal and moral culpability for overdose

Among participants, the proliferation of fentanyl in the drug supply exacerbated uncertainties around dosing and heightened feelings of moral responsibility—and even criminal liability—for overdoses experienced by those they assist. While the toxicity of the drug supply was outside of their control, many participants described feeling responsible for the immediate or later overdose deaths of those they assisted. As described by Kevin: “whenever I shoot [inject] somebody, I’m responsible for that person’s well-being. If the person OD’s that’s on me.” James stated that he had not helped inject anyone lately given that fentanyl was present in most street drugs. He further explained how the heightened potency of fentanyl-adulterated drugs increased the chances of fatal overdose, and described feeling morally culpable for the potential death of a person he assists:

When you’re injecting that person up, the feeling that you get through your body is: ‘[if] I fuck this up, this person’s gonna die…this is not good, you shouldn’t be doing this, you’re gonna kill the person.’ And then you stick the needle into the person’s body, pull back the syringe, push it back in, take the chance of killing the person. If you kill that person, all the friends of that person would jump on you, give you the baddest beating you’ve ever had from killing the person … It’s the worst feeling you got knowing a perfect stranger’s gonna die because of you if you mess up.

Alexandra similarly explains how she would have to be under considerable pressure to consider injecting someone for the first time due to the potency of fentanyl: “I am taking their life into my hands, and there’s a chance that they could pass away […] because there’s fentanyl in everything.”

Further, there was a perception that homicide laws could be applied to prosecute those who sell or otherwise provide fentanyl-adulterated substances. In one exceptional case, Sadie, a former drug courier herself, even expressed support for applying the law in this way:

And I don’t understand why they’re not caging those people that are being caught with the Fentanyl. [Yeah] If they charge, start charging people with like attempted murder and stuff like that, that would change the game drastically.

This comment is reflective of the broader political climate in North America where homicide and manslaughter laws are being considered in many jurisdictions as a way to deter people from sharing or selling opioids (L Beletsky, 2018). Within this context, assisting others with injecting was perceived by some participants as potential grounds for homicide charges. Matthew, a 54 year old white man, refused to assist strangers with injection on the grounds that he did not want to be a “murderer”:

I’m not injecting until I know what you’re doing. I’m sorry. I’m not killing you. I don’t know who you are. I’m just not. I don’t care how big of an asshole you are, murder’s not on my mind. I don’t go there. I don’t do that.

These quotes illustrate how punitive criminal justice policy and discursive constructions of fentanyl overdose as ‘murder’ come to frame how PWID understand injection assistance. That is, participants’ accounts of injection assistance suggested that they have to some extent accepted prevailing discourses in criminal justice policy that assign blame for the present overdose crisis to those who deal and deliver opioids. Of note, we could find no cases where participants directly linked fentanyl-adulteration and overdoses to drug policies or other structural determinants. The discursive framing employed by participants thus functioned to legitimize punitive criminal justice approaches to the overdose crisis (i.e. “caging those people”), while obscuring any structural forces (e.g. drug prohibition) that have directly and indirectly facilitated fentanyl-adulteration. This punitive discourse could be interpreted as a form of symbolic violence through which individuals – in this case, people initiating others into injection drug use – internalize responsibility, and subsequently moral and legal culpability, for subsequent harms (Bourdeiu and Wacquant, 1992). The emotional intensity of initiation experiences as understood within broader moral codes, and as seen in common expressions of regret and remorse about providing initiation assistance, might further amplify such feelings of culpability.

Injection assistance as a form of harm reduction during the overdose crisis

Initiation assistance was perceived by participants as morally wrong, risky, and potentially criminal. However, these concerns existed in tension with the need to protect those initiating injecting from doing so in ways that made them vulnerable to overdose or other harms. Injection initiation assistance was sometimes articulated as a responsible practice within the context of a fentanyl-driven overdose crisis. Participants expressed a common perception that once someone has set their mind to inject, it is inevitable that they will find a way to do so. Providing assistance to novice injectors was therefore seen as a form of harm reduction to prevent them from using alone or seeking assistance from someone who engages in riskier injection practices (e.g. by missing the shot or using unsterile equipment) or from someone who takes advantage of novice injectors (e.g. by stealing their drugs or purposefully dosing them too high to facilitate theft). While exploitation and violence have long been a feature of the structural vulnerability experienced by PWID in Vancouver’s Downtown Eastside, the introduction of fentanyl into the drug supply heightens risks of debilitating or fatal overdose for novice injectors. Given these circumstances, PWID such as ‘Donna’ stressed the importance of novice injectors learning how to use properly from trustworthy friends or family:

Donna: Well I did it [showed them how to inject] because the person I did it for he was kind of someone that would be kind of looked at as a mark [vulnerable or easy to exploit], you know.

Interviewer: What’s that?

Donna: It’s a person that […] you give them a little bit of extra so he OD’s or something […] With the Fentanyl going around and stuff right now […] it scared me cause I know this guy really well and he’s a good friend and I don’t want him getting sucked up by somebody else, right?

Practicing injection assistance was understood by some participants as a way of caring and being responsible for others within their social network. Particularly when it came to friends, family and intimate partners, providing injection initiation assistance aligned with participants’ role as someone who “watches over” others when they are using, and can intervene in the case of overdose:

Interviewer: How do other people feel about injecting other people for their first time? Is it looked down upon?

Vernon: It’s… well, I don’t know. It’s pretty scary. It is kind of looked down upon, but at the same time it’s kind of not, because they’re kind of, ‘Okay, we’ve got to train you because, you know, if something happens to us and we want somebody here, you know, in case we all overdose, we go down, at least somebody’s here to be with us.’

As ‘Vernon’s’ quote demonstrates, there is a tension between the moral code discouraging injection initiation (“it is kind of looked down upon”) and social norms emphasizing interpersonal responsibility to prevent harm among novice injectors (“we’ve got to train you”). Given widespread fentanyl adulteration, assisting a novice injector is therefore considered by some as a form of care to prevent fatal overdose.

Discussion

This article explores how PWID in Vancouver’s Downtown Eastside understand and navigate social norms of injection initiation within the context of an increasingly fentanyl-adulterated drug supply. Our study contributes novel evidence that the type and potency of drugs available in a market (in this case, highly potent fentanyl) can influence how PWID understand social norms of injection initiation. Our findings indicate a disjunctive and complex relationship between fentanyl adulteration and injection initiation processes. The proliferation of fentanyl may discourage injection initiation in that established injectors reported being reluctant to assist novice injectors out of concern for subsequent overdose. However, fentanyl adulteration and concerns around potency may also provide a compelling pragmatic rationale for assisting new injectors, given that established injectors expressed concern about novice injectors’ vulnerability to overdose if they use alone or receive assistance from someone less skilled.

Findings from our study help explain a discordance, observed in previous research, between PWID’s attitudes toward injection initiation and their self-reported practice. Previous research on injection assistances consistently finds that PWID report ambivalence, reluctance, and remorse about helping novices inject for the first time (Kolla, et al., 2015; Small, et al., 2009; Wenger, et al., 2016; Witteveen, Van Ameijden, & Schippers, 2006). The reluctance to assist novice injectors has been linked to group norms emphasizing the unacceptability of facilitating injection initiation (Guise, et al., 2018; Kolla, et al., 2015; Rhodes, et al., 2011; Simmons, Rajan, & McMahon, 2012; Small, et al., 2009; Wenger, et al., 2016), which have been sometimes termed a ‘moral code’(Small, et al., 2009). Despite this moral code, previous prevalence studies find that between 14 to 47% of PWID report ever providing injection initiation assistance (Bryant and Treloar, 2008; Crofts, et al., 1996; Day, Ross, Dietze, & Dolan, 2005; Fairbairn et al., 2006; Hamida et al., 2018). Our analysis suggests that ‘interpersonal responsibility’ may be a more socially meaningful and analytically useful concept to help explain why, when, and how established injectors assist novice injectors. While participants described the prohibition against injection initiation as a ‘moral code,’ decisions about whether and how to assist others for the first time were also informed by social norms around interpersonal responsibility. That is, participants measured the ethical implications of injection initiation assistance practices against a reciprocal moral duty to prevent harm to other injectors within their social network.

The notion of interpersonal responsibility articulated in interviews exist alongside, but extend beyond, concepts of individual responsibility that have come to characterize neo-liberal forms of governance in health promotion (Fraser, 2004). Critical studies in public health have documented a shift towards the individualization of responsibility that emphasizes the moral imperative for citizens to care for the self through healthy lifestyle choices (Ashton and Seymour, 1988; Rose, 1996) Harm reduction interventions have often reproduced this approach through their injunctions upon injection drug users to practice ‘responsible’ drug use by engaging in safer injection practices (Moore and Fraser, 2006). In contrast, the understanding of responsibility articulated by PWID in this study conveyed an additional responsibility for injection drug users to protect others from injection-related harms. This finding resonates with earlier research by Fraser (2004) which found that injection drug users believed they had a responsibility to protect others from hepatitis C transmission by discarding used syringes and educating younger injectors about safer injection practices. Interpersonal responsibility is analogous to the concept of ‘shared responsibility’ employed in the social sciences to describe HIV prevention messaging which appeals to consideration for others and the duty to behave in ways that protect others (Dodds, 2002; Owczarzak, 2009). While interpersonal responsibility offers an alternative to notions of individual responsibility, the power effects of these conceptualizations are similar in their tendency to lay blame for drug-related harms on the behaviours of injection drug users while overlooking the contribution of broader structural forces. This effect was evident in the tendency for participants to locate blame for fentanyl-related overdoses on other drug users who assisted with injection.

Participants’ accounts illustrate the divergent ways in which social norms around interpersonal responsibility can be interpreted and enacted. While interpersonal responsibility was in some circumstances performed by discouraging others from injecting, interpersonal responsibility could also be performed by assisting novice injectors to use in ways that limited harms associated with injecting highly potent fentanyl. The concept of interpersonal responsibility therefor allows us to extend beyond a deterministic understanding of moral codes—in which behaviour is proscribed by social norms prohibiting injection assistance—and consider the ways in which injection drug users are dynamically involved in constructing and managing social norms around injection initiation through their everyday practices and interface with broader social and economic transitions.

Structural vulnerability and the risk environment framework can also help account for why PWID in our study supported moral codes prohibiting injection initiation while also contravening such codes by engaging in the practice of initiating others (Guise, et al., 2018; Wenger, et al., 2016). While PWID exercise agency in deciding whether to assist novice injectors, this agency is constrained by social and structural conditions that limit their ability to act in accordance to group norms discouraging or prohibiting injection initiation. Previous research highlights how social norms prohibiting injection initiation prove difficult to maintain given the structural vulnerability of participants and competing priorities within the risk environment (Kolla, et al., 2015; Small, et al., 2009; Wenger, et al., 2016). Socio-economic vulnerability has been identified as a motivating factor for initiation assistance, as strong opposition to initiation of novice injectors can be overlooked due to financial opportunities to trade injection assistance for money, drugs, services, or other material resources (Fairbairn, et al., 2010; Wenger, et al., 2016).

Given the structural circumstances PWID face—in this case, an increasingly potent and unpredictable supply of fentanyl-adulterated drugs—providing injection assistance may be a way of practicing interpersonal responsibility toward novice injectors within drug use networks, with the goal of reducing potential harms such as overdose. Within this context, initiation assistance may even be interpreted as an informal supervised injection service that PWID practice to reduce overdose risk amongst novice injectors. This interpretation is supported by previous qualitative studies in which PWID describe assisting as a form of harm reduction to prevent possible harms inflicted by novice injector’s own attempts (Guise, et al., 2018; Kolla, et al., 2015; Wenger, et al., 2016). Our study also found that some established injectors perceive there will be negative consequences if a novice injector is unable to convince a skilled and trusted injector to assist them (e.g. missed injection from injecting themselves, or being taken advantage of by strangers). These concerns are not unfounded, as previous research finds that requiring help injecting and recent transition to injection drug use are associated with elevated risk of overdose and blood-borne disease transmission among PWID (Fennema, Van Ameijden, Van Den Hoek, & Coutinho, 1997; T. Kerr et al., 2007; Miller et al., 2002; O’Connell et al., 2005). Injection assistance, with attention to dosing and tolerance, may help mediate risk of overdose amongst less experienced injectors(T Kerr, Oleson, Tyndall, J, & Wood, 2005; McNeil, Small, Lampkin, Shannon, & Kerr, 2014; W Small et al., 2012). Therefore, efforts to prevent transitions to injection drug use must be balanced with the need to ensure novice injectors also have access to hygienic and safer injecting environments (e.g. supervised injections sites) that reduce injectors’ risk of fatal overdose and other injection-related harms. There are currently few sanctioned supervised injection sites in North America where PWID may receive assistance with injections, which poses a structural barrier to overdose risk reduction for novice injectors and others who require assistance to inject (McNeil, et al., 2014). The implementation and geographic scale-up supervised consumption services that provide injection assistance (currently being trialed in three Canadian supervised injection sites) may alleviate some of the concerns that established PWID have about novice injectors injecting themselves (Government of Canada, 2018; McNeil, et al., 2014).

Our finding that participants were concerned about being legally and morally responsible for overdose deaths of those they assist is likely connected to the emergence of laws in North America since the 1980s that allow prosecutors to pursue manslaughter or homicide charges against people who sell or provide drugs linked to a fatal overdose (Beletsky, 2018; Drug Policy Alliance, 2017). These laws have gained renewed attention from prosecutors in the US and Canada as legal tools to address the overdose crisis, despite the lack of evidence that they deter injection drug use (Beletsky, 2018). While police in Vancouver have not yet laid manslaughter or homicide charges against those ‘trafficking’ fentanyl (including those who assist with injection), perceptions that the law is applied in this way, or may be in the future, appears to be influencing norms around injection initiation and other forms of peer-to-peer assistance, with potentially adverse consequences. Fear of legal culpability may heighten risk of fatal overdose if it discourages those who assist from seeking help during an overdose. While we found no cases of this scenario in our interviews, the desire to appear responsible in the interview could plausibly dissuade participants from disclosing this practice. Further, various studies from the United States—where drug-induced homicide prosecutions have been pursued more aggressively—find that fear of police involvement is the most common reason why people avoid calling 911 during an overdose (Davidson, Ochoa, Hahn, Evans, & Moss, 2002; Tracy et al., 2005). The growing implementation of Good Samaritan Laws, which provide limited immunity from drug-related offences for those calling for help during an overdose, have the potential to reduce barriers to seeking medical assistance during an overdose (Banta-Green, Kuszler, Coffin, & Schoeppe, 2011). However, the potential of Good Samaritan laws are undermined by drug homicide laws which may discourage people from seeking help for fear of prosecution for murder or manslaughter (Drug Policy Alliance, 2017). Such approaches should be reconsidered given their potential to produce disproportionate harms among PWID.

Finally, it is worth considering the macro-level structural forces that have facilitated the proliferation of fentanyl within Vancouver’s drug supply. A punitive and supply-interdiction approach to drug policy has arguably contributed to the present influx of fentanyl within the drug supply by creating an economic incentive to traffic more concentrated opiates that are more difficult to detect (Beletsky and Davis, 2017). At the same time, there is evidence that sales for street opioids have increased as access to licit opioids (e.g. OxyContin) have been reduced due to concerns about over-prescribing (Martin, Cunliffe, Decary-Hetu, & Aldridge, 2018). Policy decisions to interdict the supply of safer opioids interact with the socio-economic marginalization of poor PWID to create conditions of structural vulnerability in which PWID have limited ability to reduce their risk of fentanyl-related overdose. Efforts to disrupt injection initiation and reduce injection-related harms must consider these broader structural forces limiting PWID’s access to a safe supply of opioids. While participants were hesitant to assist novice injectors due to the toxicity of fentanyl and fentanyl-adulterated opioids, many ultimately chose to do so as a form of harm reduction. This finding casts further doubts on the effectiveness of drug prohibition and supply-interdiction policies to deter injection initiation. Rather, it is possible that PWID would not experience the same pressures to transition into illicit injection or assist novice injectors as a form of harm reduction if there was a more consistent supply of opioids free of fentanyl. This may be achieved by increasing access to opioid agonist programs (e.g., heroin assisted treatment, hydromorphone, buprenorphine) and enacting policies conductive to a safer supply of opioids, including those that end drug prohibition and facilitate distribution of pharmaceutical-grade opioids.

This study has several limitations. While in-depth interviewing allowed us to explore PWID’s experiences and perceptions of injection initiation assistance, we were not able to triangulate interview data with naturalistic observation of participants’ injection assistance practices, which would provide better insight into how participant’s perceptions and accounts align with their everyday practices. Second, we acknowledge that the context of the interview itself influenced the construction of narratives about injection initiation. Since the interviews were conducted within a public health research setting, we assume that participants emphasized views and practices that aligned with public health principles of responsible practice. This desire to be perceived as responsible and ethical may explain why pragmatic reasons for facilitating injection (e.g. in exchange for drugs, sex and money) did not feature prominently in participants’ narratives. Third, although many participants had previously experienced homelessness at some point in their lives, all were in some form of housing at the time of interviews. As such, the findings reported here may not be representative of those who experience heightened structural vulnerability from homelessness. Finally, our findings provide insight into localized injection initiation norms within a cohort of PWID residing primarily in Vancouver’s concentrated downtown drug use scene. While these findings may not be generalizable to other settings, our findings about injection assistance reported here are consistent with those reported in previous research from other settings (Guise, et al., 2018; Kolla, et al., 2015; Wenger, et al., 2016).

In conclusion, the proliferation of fentanyl produces complex ethical dilemmas for marginalized PWID, who must weigh potential harms of initiation assistance against the desire to reduce novice injectors’ overdose risks. Fentanyl adulteration in the drug supply, and the related rise in overdose deaths, means that injection initiation assistance takes on shifting and sometimes conflicting meanings for PWID: as a morally prohibited practice, a potentially criminal activity, but also as a form of harm reduction. Within this context, PWID are forging new strategies to exercise interpersonal responsibility, which at times may encompass assisting novice injectors who are at heightened risk of harm due to their lowered tolerance, limited experience with risk reduction practices, and vulnerability to being “taken advantage of” within drug use scenes. Fears of culpability for overdoses may amplify risk of overdose deaths if this discourages PWID from seeking help during an overdose. These findings draw attention to the limitations of individual-focused behaviour change interventions to prevent injection initiation, and suggest that structural interventions (e.g. supervised injection sites allowing for assisted injection, Good Samaritan laws, and policy changes conducive to a safer drug supply) should be prioritized to reduce the vulnerability of novice injectors to overdose and other harms.