Introduction

Injecting drug use has serious implications with regards to transmission risk of blood borne viruses (BBV) such as HIV and hepatitis C. Because many seroconversions happen within the first few years of injecting (van Beek et al. 1998) understanding the practices and contexts that constitute the experience of initiation to injecting is important. Available information about initiation to injecting tends to focus on the experience of the initiate and not the initiator. This research reveals how initiates tend to have socio-demographic profiles characterised by disadvantage where they are likely to be unemployed and/or homeless (Neaigus et al. 2006) after having left school early and/or having been incarcerated (Crofts et al. 1996). Initiates are also likely to have experienced recent abuse and violence (Fuller et al. 2002; Neaigus et al. 2006). Importantly, initiation to injecting appears to be strongly mediated through relationships with others where the first injection is most commonly facilitated by a friend or sexual partner (Bryant and Treloar 2007; Day et al. 2005; Diaz et al. 2002; Doherty et al. 2000; Roy et al. 2002) who is likely to be an older and longer-term injecting drug user (IDU) (Day et al. 2005; Fuller et al. 2002; Roy et al. 2002). The importance of relationships in initiation makes it necessary to examine the experiences of initiators in order to provide a full understanding. However, there is a lack of research literature about the experiences, knowledge and practices of initiators. A related area to draw upon in this regard relates to the risk practices of those who assist others to inject but do not specifically initiate others to injecting. Those who provide injecting assistance are more likely to borrow used syringes from others (Kral et al. 1999), lend used syringes to others (Fairbairn et al. 2006; Kral et al. 1999) and share peripheral equipment such as cookers and filters (Kral et al. 1999). Friedman et al. (1998) show how those who inject others (known as ‘hit doctors’ or ‘street doctors’ in the United States) also tend to be deeply embedded within drug using networks, revealing the way that injecting others is linked to having close relationships with other drug users. Importantly, even though ‘hit doctors’ seem to be involved in activities that pose a high risk of transmission, Friedman et al. (1998) suggest that their embeddedness within networks means that “their actions can help to prevent the spread of epidemic diseases” (Friedman et al. 1998, p. 1414). Thus, examining the harm reduction practices and knowledge of initiators seems important, in particular the degree to which they pass on information to initiates.

In sum, understanding the context and practices of initiation to injecting drug use must include an understanding of the initiation experience from the viewpoint of the initiate and initiator. This paper uses data collected from early-career IDU to describe the socio-demographic characteristics, drug and injecting practices, and harm reduction knowledge and practices of IDUs who report having initiated others to injecting.

Method

Participants

The study population was recruited as a part of the Initiations and Transitions to Drug Use Study, which sought to examine young injectors’ experiences at initiation to injecting (see Abelson et al. 2006; Bryant and Treloar 2007; Treloar and Abelson 2005). Eligibility criteria included: (1) current or recent injecting drug use (within previous 6 months), (2) injecting history of 5 years or less (to reduce recall bias), and (3) aged 16–25 years. Participants were recruited by convenience sampling between December 2000 and February 2002 from two urban and one regional location in two Australian states (New South Wales and Queensland). Recruitment fliers were posted in settings that young IDU were likely to attend such as youth shelters, needle and syringe programs, public health clinics, emergency rooms, and treatment centres. Participants were also recruited using snowball sampling whereby participants who had completed interviews were asked to pass on study information and contact details to other young IDU.

A structured questionnaire was administered in a face-to-face interview. Most interviews (84%) were conducted by peer interviewers, the remainder by a research assistant. Peer interviewers were current IDU aged 16–25 years who received research training in eligibility screening, consent procedures and interview administration. Peer involvement has been shown to be an effective means of accessing hidden populations of IDUs, including newly initiated and young users (Williams and Roche 1999). Peer interviewers facilitate recruitment because of a sense of shared norms, values and meanings associated with drug use (Kelsall and Kerger 2001; Williams and Roche 1999). Each participant received $20AU for completing an interview.

Measures

The questionnaire was developed from extensive literature review and consultation with key informants. Data were grouped according to two questions relating to initiating others: ‘Have you ever given anyone their very first injection of anything?’ and ‘How many people have you given their first injection to?’ The questions defined initiating others as the act of injecting somebody else and not ‘teaching’ or ‘showing’ others, the meaning of which can be subjective. Five groups of independent variables were selected based on known associations evident in previous research and items of interest in this analysis. Five socio-demographic variables were included: age at time of interview (years), gender (male vs. female), income (unemployment benefits vs. other), Aboriginality (yes vs. no), and been in prison or detention recently (≤12 months) (yes vs. no). Four drug use variables were included: years of injecting, recent polydrug use (≤1 month) (yes vs. no) defined as injecting two or more illicit drugs not necessarily at the same time, frequency of injecting (≤1 month) (not last month, once a week or less, more than once a week but less than everyday, everyday or more often), and last drug injected (heroin/methadone, speed/cocaine, other). Four variables described recent injecting practices (≤6 months): borrowed used syringes (yes vs. no), lent used syringes (yes vs. no), injected others (yes vs. no), and had been injected by others (yes vs. no). Seven variables described knowledge and practice of harm reduction: four items assessing knowledge about BBV transmission, two items about source of sterile needles and syringes (formal sources such as needle and syringe programs and pharmacies; informal sources such as friends and dealers) and one item about telling others about hepatitis C and safe injecting (yes vs. no). Two variables regarding BBV testing and status: having been recently tested for hepatitis C (≤12 months) (yes vs. no), self report of being hepatitis C positive (yes vs. no). (Data were collected about HIV testing and status however only two participants reported being HIV positive, which is congruent with epidemiological estimates showing the low prevalence of HIV among Australian IDUs (National Centre in HIV Epidemiology and Clinical Research 2007)).

Data Analysis

We wished to calculate the total number of other people initiated by initiators in the study. This was done by multiplying the number of initiators by their reported number of people initiated. Odds ratios (OR) and 95% confidence intervals (CI) comparing initiators to non-initiators were estimated using logistic regression. Data were checked to ensure their suitability for logistic regression by examining expected cell counts using cross tabulations (Tabachnick and Fidel 1996). OR were adjusted for years of injecting and any potential differences created by the sampling strategy (recruitment location, peer vs. non-peer interview, and Aboriginality since Aboriginal people were over represented in the sample).

Results

In total, 336 participants were recruited. Twelve participants did not provide data about whether they had initiated others and were excluded, leaving 324. Almost half (49%) of the participants were recruited from Sydney, 35% from Brisbane, and 16% from the regional location. A total of 55 participants (17%) reported that they had initiated someone else to injecting (Table 1). Of these, almost half (n = 24, 43.6%) reported they had initiated only one other person, 12.7% (n = 7) reported initiating two others, 9% (n = 5) initiated three others, and the remainder of initiators (20%, n = 11) reporting that they had initiated four or more other people (Table 1). The 55 initiators reported initiating a total of 128 others (Table 1), or about 2.3 people each. Most commonly, initiates were friends of the initiator (n = 35, 63.6%) or a sexual partner (n = 17, 30.9%). The proportion of participants reporting that they initiated others rose in relation to the number of years they had been injecting, from 8.5% of those within the first year, to 20.0% of those who had been injecting 2–3 years, to 50.0% of those injecting for 4–5 years (χ2 = 26.36, 4 df, p < 0.001) (Table 1). As expected, initiators had been injecting for a significantly longer period than non-initiators (3.2 vs. 2.4 years; OR 1.85, 95%CI: 1.38–2.48) (Table 2).

Table 1 Number of people initiated by initiators, by years of initiators’ injecting
Full size table
Table 2 OR and 95%CI for initiators compared to non-initiators
Full size table

ORs indicate that, compared to non-initiators, initiators were more likely to have been in prison or detention recently (OR 2.92, 95%CI: 1.40–6.11), to have lent a used syringe (OR 2.36, 95%CI: 1.19–4.67) and injected others recently (OR 7.47, 95%CI: 3.62–15.43), to have obtained needles and syringes from informal sources such as friends or dealers (OR 3.28, 95%CI: 1.72–6.24), to have told others about hepatitis C and safe injecting (OR 2.36, 95%CI: 1.26–4.40), and to have been recently tested for hepatitis C (OR 2.44, 95%CI: 1.26–4.71) (Table 2). ORs also show that initiators were less likely to have been injected by others recently (OR 0.44, 95%CI: 0.20–0.97) (Table 2).

Discussion

This analysis reveals the considerable proportion of young, early-career IDU who report giving others their first injection. Within the first 5 years of their own injecting, each initiator in this study had, on average, given two other people a first injection. Thus, of the total 324 young injectors in the sample, another 128 young people were introduced to injecting over a 5-year period, through the actions of 55 initiators. This shows how the act of injecting can itself be understood “as an ‘infectious’ or at least ‘communicable’ process” (Crofts et al. 1996, p. 1194), albeit one that is underpinned by complex social processes. Previous research indicates that some initiates actively seek to be initiated to injecting (Crofts et al. 1996), suggesting that at least some the 128 people who were initiated in this study might have started injecting regardless of the role of their initiator. Certainly, social research shows how people have agency in their drug use, in particular within sexual relationships where engaging in drug use as a couple has been shown to create meanings of intimacy (Rhodes and Quirk 1998). Here, a sexual partner may actively seek to be injected in order to imbue a relationship with feelings of emotional closeness. Our data indicate that a considerable proportion (30%) of initiators were in a sexual relationship with their initiate, suggesting that the ‘reproduction’ of the injecting population in our study may have complex social underpinnings.

The paper reveals significant differences between initiators and non-initiators, particularly in relation to harm reduction knowledge and practices. Initiators were more likely than non-initiators to report that they told others about hepatitis C and safe injecting. This suggests that some initiators may be well-placed to provide first injections to others if they are also more likely to teach them about BBV transmission. However, we did not collect data about the content and quality of the information initiators passed on. Previous research identifies that the information passed through networks of IDU can be poor quality (Southgate and Hopwood 2001; Treloar and Abelson 2005). The data in this study shows that initiators did not have more accurate knowledge about hepatitis C than non-initiators, and that both groups had poor knowledge about some aspects of hepatitis C (that there is more than one type and that a person can be infected more than once). Moreover, while initiators reported that they commonly used formal sources of needles and syringes (98% had used a needle and syringe program or pharmacy in the last 6 months), they were more likely than non-initiators to have also obtained needles and syringes from informal sources such as friends and dealers where the sterility of needles cannot be guaranteed. Thus, while initiators might be a potential mode of preventing initiation to injecting (since they can act as agents in discouraging requests for initiation) (Friedman et al. 1998; Hunt et al. 1999; Shelley et al. 1993), the findings in this paper suggest that, even though they were more likely to pass on information, they did not have more accurate knowledge of BBVs, and they commonly obtain needles from sources that were potentially unsafe.

The generalisability of the study data are limited because we used a self-selected convenience sample. In many situations, this sampling method is the most suitable way to collect information from IDU (Kaye and Darke 2004) but it means that the study findings may not be generalisable. Moreover, the study findings are based on self-report, which can be problematic when relating to stigmatised practices such as injecting drug use (Latkin et al. 1993; Latkin and Vlahov 1998). However, others have found self-report data from drug users about their drug use patterns to be reliable (Darke 1998). We believe using peer interviewers can reduce the possibility of biases resulting from the self-report of stigmatised practices because drug users are likely to be more candid with their peers (Kelsall and Kerger 2001). Finally, the study data are likely affected to some extent by recall bias of participants attempting to remember an event such as the initiation of another user that may have happened up to 5 years ago.

In sum, this analysis reveals the importance of examining the experiences of initiators and not just initiates. It shows that young, early-career IDU commonly initiate others to injecting, but may not be well-positioned to do so. While they are more likely to pass on harm reduction information, we do not know the content and quality of this information. Initiators in this study did not have more accurate knowledge of BBV, and commonly obtained needles and syringes from sources that were potentially unsafe. Intervention strategies aimed at reducing initiation to injecting should include young users and might also consider supporting them to increase their knowledge and practices relating to harm reduction so that this can be passed on to initiates.