Abstract
Introduction and Aims:
People who inject drugs (PWID) play critical roles in assisting others into injection drug use (IDU) initiation. Understanding perceptions of PWID’s risk of initiating others is needed to inform interventions for prevention. The objective was to examine factors associated with self-perception of assisting with future IDU initiation events. The primary variables of interest are the relationships of PWID with the person(s) they assisted and their reasons for previously providing initiation assistance.
Design and Methods:
Data from Preventing Injecting by Modifying Existing Responses, a multi-site prospective community-recruited cohort study, were analysed. Analyses were restricted to PWID who reported ever providing IDU initiation assistance. Site-specific (Vancouver, Canada [n = 746]; San Diego, USA [n = 95] and Tijuana, Mexico [n = 92]) multivariable logistic regression analyses were performed to determine factors associated with self-perception of assisting with future IDU initiation.
Results:
Having provided IDU initiation assistance to a family member or intimate partner decreased the odds of self-perception of assisting with future IDU initiation in Vancouver (AOR = 0.4; 95% CI 0.2–0.8); however, previous IDU initiation assistance to an ‘other’ increased the odds of self-perception of assisting with future IDU in Tijuana (AOR = 12.0; 95% CI 2.1–70.3). Daily IDU (Vancouver: AOR = 3.7; 95% CI 2.1–6.4) and less than daily IDU (San Diego: AOR = 5.9; 95% CI .3–27.1) (Vancouver: AOR = 2.6; 95% CI 1.4–2.9) were associated with increased self-perception of assisting with future IDU compared to current non-injectors.
Discussion and Conclusions:
Relationship to past initiates and IDU frequency might increase PWID’s self-perception of assisting with future IDU. Interventions focused on social support and reducing IDU frequency may decrease occurrences of IDU initiation assistance.
Keywords: North America, drug injecting initiation, injection drug use, injection initiation assistance, relationships
INTRODUCTION
Injection drug use is widespread globally, with recent estimates suggesting that there are 15.6 million people who inject drugs (PWID) aged 15–64 years worldwide [1]. Drug injection is a major source of morbidity (e.g. non-fatal overdose) and mortality, and contributes to HIV and hepatitis C transmission [2, 3]. Preventing transitions into drug injection is critical to prevent a range of downstream harms [4, 5]. To that end, a growing body of research has characterised risk factors for injection initiation, including the contexts, circumstances and experiences of first-time injectors (e.g. experiencing homelessness, having friends or peers who inject drugs, residing in neighbourhoods with endemic drug presence) [6–9].
Research has increasingly focused on the role of PWID in facilitating the initiation of others into injection drug use [5, 9, 10], with evidence indicating that an individual’s relationships—including those with family, friends, intimate partners and acquaintances—can influence a person’s decision to assist others into injection drug use [7, 11–14]. In particular, among injection-naïve persons, exposure to injection drug use through one’s social network is associated with requesting injection initiation assistance from someone who is experienced with drug injecting [14–17]. Further, data suggest that many individuals are assisted into injection drug use by someone they know [10, 18]. While those seeking to initiate are often the source of requests for initiation [11, 17, 19], injection drug use initiation may occur within the context of peer pressure or coercion, specifically among injection-naïve women [11]. One study focused on PWID in California found that among those initiated, 42% were initiated by acquaintances, 37% by friends, 4% by intimate partners and 2% by family members [19].
While the majority of PWID report refusing to assist others into injecting (73% of PWID in one sample reported refusing to initiate others), those who accept may assist multiple people [9, 15]. For example, in a 2008 study of 324 PWID in Australia, 55 (17%) indicated they had assisted in another person’s injection initiation; this subsample was implicated in a total of 128 injection drug use initiations within the first 5 years of their own initiation into drug injection [15, 19]. A 2016–2017 study in Los Angeles and San Francisco, California, USA, found 132 PWID had provided injection initiation assistance to 784 people in the prior 6 months, with 34 PWID accounting for 84% of the total number of people initiated [20].
This paper draws upon multiple conceptual models to contextualise the processes by which injection practices are communicated among communities of people who use drugs. The Socio-Ecological Model (SEM) highlights the complex interplay between individual, relationship, community and societal factors that, in concert, influence how an individual interacts with the people, places, institutions and policies surrounding them [21, 22]. The SEM proposes that multiple environmental (e.g. policy, health systems) and relational (e.g. friend, family and romantic relationships, peers) influences contribute to individual behaviour [23]. It is often used to highlight an individual’s risk (e.g. marginalisation, stigma, limited resources) and protective factors (e.g. safety, social connectedness) [24] and to inform intervention strategies that target multiple entry points (e.g. individual, community, society-level) [25]. In addition, we incorporated the social learning model of injection initiation developed by Strike and colleagues [26]. This approach delineates multiple pathways by which injection knowledge and interest in injection initiation may be communicated through a social network: for example, increased interest through PWID speaking positively about injecting, injecting in front of non-injectors, guiding the injection or directly assisting in injecting (e.g. didactic and interpersonal influences) or decreased interest in injection initiation through engaging in treatment and behavioural interventions focused on peer-to-peer accountability (e.g. intrapersonal influences).
Taking that into account, for the purposes of the present study, we employed the SEM and social learning model to understand the association between PWID’s relationship to those they have provided injection drug use assistance to in the past and their perceived likelihood of initiating someone into injection drug use again in the future. In particular, we used these frameworks to investigate the social contexts in which individuals who have assisted injection-naïve persons into injection drug use interact with others, and to identify factors that may be responsive to prevention efforts among this population [27].
Despite a growing evidence base on the role of PWID in facilitating the process of injection initiation, little research has focused on the reasons why PWID may assist others into injection drug use. Understanding PWID’s motives for providing injection drug use initiation assistance and their self-perception of assisting with future injection drug use initiation can inform injection drug use prevention and harm reduction efforts [11]. We therefore sought to identify factors associated with a self-perception of assisting with future initiation among PWID who have previously participated in the initiation of others.
METHODS
Preventing Injecting by Modifying Existing Responses (PRIMER) is a multi-site mixed methods study pooling data on injection drug use initiation assistance and other factors from cohort studies of PWID in North America [28]. Details regarding PRIMER study methods have been previously described in full [28]. Briefly, PRIMER employs data from prospective community-recruited cohorts of PWID to assess how socio-structural factors may influence the risk that PWID provide injection initiation assistance to others. Open enrolment was undertaken through extensive street-based outreach by key staff members and peers. Participants were eligible if they reported recent (i.e. last 6 month) injection drug use at baseline.
For the present study and analysis, quantitative data from the three North American cohort studies participating in PRIMER were used: the Proyecto El Cuete IV (ECIV) in Tijuana, Mexico [29, 30], the Study of Tuberculosis, AIDS, and Hepatitis C Risk (STAHR II) in San Diego, USA [30], and the linked Vancouver Drug User Study and AIDS Care Cohort to Evaluate Exposure to Survival Services (VDUS/ACCESS) in Vancouver, Canada [31–34]. Participants provided written informed consent. This study was approved by the University of California San Diego Human Research Protection Program (UCSD IRB#150866), the University of British Columbia-Province Health Care Research Ethics Board and the Universidad de Xochicalco Ethics Committee.
Survey questionnaires are broadly comparable, and ECIV and STAHR were specifically designed as linked binational studies [30]; these surveys were also constructed based in part on the original iteration of the VDUS/ACCESS survey [28]. Identical survey items assessing: (i) PWID’s experiences with injection drug use initiation assistance; and (ii) PWID’s self-perception of assisting with future injection drug use initiation, were introduced into participating cohort questionnaires. This was done in September 2014 and coincided with follow-up 7 in ECIV, follow-up 4 in STAHR II and follow-up 18 in VDUS/ACCESS. Follow-ups were conducted every 6 months at each site. By the time of the analysis, we had five waves of data collection from VDUS/ACCESS that included the initiation experience questions; seven waves of data collection from ECIV; and one wave of STAHR II data (due to funding end). The current analysis selected the visit where a subject first reported having ever provided injection drug use initiation assistance and responded to the future initiation question, approximately 30.0% (n = 933) of the total PRIMER sample (n = 3113). In Vancouver, 747 participants reported they had ‘ever initiated’ someone and 746 responded to the future initiation question (37.0% of the Vancouver sample, n = 2017). In San Diego, 134 participants reported they had ‘ever initiated’ someone and 95 responded to the future initiation question (27.3% of the San Diego sample, n = 348). In Tijuana, 270 participants reported they had ‘ever initiated’ someone and 92 answered the future initiation question (12.3% of the Tijuana sample, n = 748).
The primary outcome of the analysis was self-perceived likelihood of assisting in an injection drug use initiation event in the future and was assessed by the following survey item: ‘How likely is it that you would initiate someone into injection drug use in the future?’; response categories were ‘definitely’, ‘probably would’, ‘definitely would not’, ‘probably would not’ and ‘unsure’. For the purposes of the analysis, responses were dichotomised into ‘yes’ (e.g. ‘definitely would’, ‘probably would’) and ‘no’ (e.g. ‘definitely would not’, ‘probably would not’). Those who responded ‘unsure’ (n = 39) were excluded from the analysis.
We identified two primary independent variables of interest: (i) the relationship(s) a participant had to the person(s) they had assisted in injection drug use initiation; and (ii) the rationale a participant provided for having assisted with injection drug use initiation.
The relationship variable captured any relationship(s) a participant had to those they had assisted in injection drug use initiation within their lifetime. For example, during their lifetime, a participant could have assisted in the injection drug use initiation of multiple people, such as their mother or boyfriend. The variables used to assess the relationship between participant and initiate were based on the following survey item: ‘What was your relationship to the people you helped inject? (select all that apply)’. We restructured all potential survey item responses into three categories: family member/intimate partner (e.g. sibling, child, girlfriend, boyfriend, sex partner), friend (e.g. friend, acquaintance) or other (e.g. stranger, inmate, other). We then derived a relationship variable with the following four mutually exclusive categories: family member or intimate partner, friend, other, more than one relationship type.
We undertook a similar approach to construct a categorical variable for injection drug use initiation rationale, which was determined by the following survey item: ‘In the past 6 months, when you helped someone inject who had never injected before, what were the reasons?’ and was coded as follows: Personal reasons (e.g. reasons that benefited the subject such as payment, being offered to share the drugs or wanting to share the experience, being offered sex or requesting a favour in return), interpersonal reasons benefitting the first-time injector (e.g. the person providing initiation assistance indicated they were worried about the injection-naïve person’s safety, the injection-naïve person(s) were not honest and indicated they had previously injected, they could not find a vein, they were too sick or otherwise incapacitated to inject themselves, they asked for help, or they did not know how), or more than one reason (i.e. selected more than one of the above categories). Some participants provided an ‘other’ response, which was re-categorised into one of the aforementioned categories (e.g. personal, interpersonal or more than one reason) (see Table S1).
We also included the following potential covariates of interest: age, sex and frequency of injection drug use in the last 6 months. Injection drug use frequency was defined categorically as ‘none’, ‘less than daily’ and ‘daily’.
Participant characteristics were summarised by cohort and overall. Comparisons among cohorts used chi-square tests for the categorical variables and analysis of variance tests for the continuous variables. Factors associated with self-perception of assisting with future injection drug use initiation were assessed by using cross-tabulation and Fisher’s exact test, separately for each cohort. Variables significant at the P < 0.05 level in bivariate analyses at any individual site were included in a multivariable logistic regression to assess whether type of relationship with past initiates was associated with self-perception of assisting with future injection drug use initiation, adjusting for age, sex and injecting frequency in the past 6 months. Due to zero-cells occurring in the crosstabs, a bias-reduced logistic regression model was performed, which addresses the issues that small-cell or zero-cell responses can introduce into the analysis (e.g. data are spread too thinly across too many cells, leading to insufficient certainty to allow for reliable statistical conclusions) [35], and thereby ensures that findings are exempt from small-sample bias (i.e. an over- or under-estimation of the parameters of the population) [36]. Statistical analyses were performed in R (version 3.5.2).
RESULTS
Participant characteristics (total n = 933) are summarised in Table 1 (El Cuete IV n = 92; STAHR II n = 95, VDUS/ACCESS n = 746). Most participants were male, with a mean age of 39 years in Tijuana (SD 9.9), 41 years in San Diego (SD 12.3) and 43 years in Vancouver (SD 12.9) (P < 0.05). In Tijuana, the majority of participants reported injecting daily (90.2%) in the past 6 months, whereas in San Diego and Vancouver, less than half of participants reported daily injection (28.4% and 37.5%, respectively, P < 0.01).
Table 1.
Sociodemographic characteristics and injection risk behaviours by cohort among people who inject drugs who have provided previous injection drug initiation assistance in Tijuana, San Diego and Vancouver (PRIMER; n = 933), n (%)
| SITES | |||
|---|---|---|---|
| Tijuana, Mexico (n = 92) | San Diego, USA (n = 95) | Vancouver, Canada (n = 746) | |
| Age, years, mean (SD) | 39.3 (9.9) | 41.0 (12.3) | 43.1 (12.9) |
| Sexa | |||
| Men | 61 (66.3) | 67 (71.3) | 488 (65.5) |
| Women | 31 (33.7) | 27 (28.7) | 256 (34.4) |
| Drug injection frequency (6 months) | |||
| None | 6 (6.5) | 32 (33.7) | 240 (32.2) |
| Less than daily | 3 (3.3) | 36 (37.9) | 226 (30.3) |
| Daily | 83 (90.2) | 27 (28.4) | 280 (37.5) |
| Relationship to initiate | |||
| Family member/intimate partner | 1(1.1) | 11 (11.6) | 86 (11.8) |
| Friend | 71 (77.2) | 52 (54.7) | 491 (67.1) |
| Other | 9 (9.8) | 6 (6.3) | 53 (7.2) |
| More than one relationship | 11 (12.0) | 26 (27.4) | 102 (13.9) |
| Reason why assisted others into injection drug use | |||
| Personal reason | 15 (20.8) | 5 (5.9) | 46 (6.7) |
| Interpersonal reason | 45 (62.5) | 31 (36.0) | 575 (83.3) |
| More than one reason | 12 (16.7) | 50 (58.1) | 69 (10.0) |
| Would inject someone in the future | |||
| Definitely would | 5 (5.4) | 6 (6.3) | 46 (6.2) |
| Probably would | 19 (20.7) | 8 (8.4) | 79 (10.6) |
| Unsure | 19 (20.7) | 9 (9.5) | 39 (5.2) |
| Probably would not | 19 (20.7) | 14 (14.7) | 136 (18.2) |
| Definitely would not | 30 (32.6) | 58 (61.1) | 446 (59.8) |
Column percentages add up to 100.
For sex in San Diego, this number also includes three transgender respondents.
PRIMER, Preventing Injecting by Modifying Existing Responses.
When participants were asked who they initiated in the past, the majority responded ‘friends/acquaintances’ across all sites (77.2% in Tijuana, 54.7% in San Diego and 67.1% in Vancouver) and provided an interpersonal reason (e.g. the person providing initiation assistance indicated they were worried about the injection-naïve person’s safety) for why they helped them initiate (62.5%, 36.0% and 83.3%, respectively). Only 1.1% of participants had provided initiation assistance to an intimate partner or family member in Tijuana, compared to almost 12% in San Diego and Vancouver. When asked if they would assist someone in initiating injection drug use in the future, a higher proportion of participants in Tijuana indicated they ‘definitely’ or ‘probably’ would compared to other sites (26.1% in Tijuana, 14.7% in San Diego,16.8% in Vancouver, P < 0.01).
Table 2 displays the bivariate associations with self-perception of assisting with future injection drug use initiation. The type of relationship with the persons they assisted was significant in Tijuana (P = 0.01), as 77.8% of those who had provided initiation assistance to someone in the ‘other’ category in the past reported a self-perception of assisting with future injection drug use initiation compared to 26.8% of those who had provided initiation assistance to friends. In Vancouver, the type of relationship with the persons assisted was marginally insignificant (P = 0.06) and not significant in San Diego (P = 0.36). In Tijuana, only one case reported assisting a family member or intimate partner; consequently, that category was removed in the multivariable analysis. Reasons for assisting someone with injection drug use was not significantly associated with self-perception that they will assist with future injection drug use initiation in any of the sites. Lastly, injection frequency in the last 6 months was significantly associated with self-perception of assisting with future injection drug use initiation in San Diego (P < 0.05) and Vancouver (P < 0.001), but not in Tijuana (P > 0.99).
Table 2.
Bivariate association with self-perception of assisting with future injection drug use initiation among who have provided previous injection drug initiation assistance by site (PRIMER; n = 865), n (%)
| Tijuana, Mexico (n = 73) | San Diego, USA (n = 86) | Vancouver, Canada (n = 707) | |||||||
|---|---|---|---|---|---|---|---|---|---|
| Would Initiate (n = 24) | Would not initiate (n = 49) | P | Would Initiate (n = 14) | Would not initiate (n = 72) | P | Would Initiate (n = 125) | Would not initiate (n = 582) | P | |
| Age, years, mean (SD) | 40.9 (9.5) | 38.3 (11.2) | 0.50 | 50.1 (14.1) | 43.6 (12.1) | 0.09 | 40 (13) | 43.9 (12.9) | <0.01 |
| Sex | |||||||||
| Men | 18 (36.7) | 31 (63.3) | 0.43 | 11 (18.6) | 48 (81.4) | 0.85 | 94 (20.3) | 370 (79.7) | 0.02 |
| Women | 6 (25.0) | 18 (75.0) | 3 (13.0) | 20 (87.0) | 31 (12.5) | 210 (87.5) | |||
| Transgender | - | - | 0 (0) | 3 (100) | - | - | |||
| Drug injection frequency, last 6 months | |||||||||
| None | 1 (20.0) | 4 (80.0) | >0.99 | 2 (6.3) | 30 (93.7) | <0.05 | 19 (8.1) | 216 (91.9) | <0.001 |
| Less than daily | 1 (33.3) | 2 (66.7) | 9 (30.0) | 21 (70.0) | 41 (18.9) | 176 (81.1) | |||
| Daily | 22 (33.9) | 43 (66.1) | 3 (12.5) | 21 (87.5) | 65 (25.5) | 190 (74.5) | |||
| Relationship to past initiates | |||||||||
| Family member/intimate partner | 0 (0) | 1 (100) | 0.01 | 0 (0) | 10 (100) | 0.36 | 7 (8.2) | 78 (91.8) | 0.06 |
| Friend | 15 (26.8) | 41 (73.2) | 9 (19.2) | 38 (80.9) | 87 (18.6) | 381 (81.4) | |||
| Other | 7 (77.8) | 2 (22.2) | 0 (0) | 6 (100) | 11 (21.6) | 40 (78.4) | |||
| More than one relationship | 2 (28.6) | 5 (71.4) | 5 (21.7) | 18 (78.3) | 19 (20.9) | 72 (79.1) | |||
| Reason for initiating | |||||||||
| Personal reason | 4 (26.7) | 11 (73.3) | 0.67 | 2 (40.0) | 3 (60.0) | 0.29 | 7 (15.2) | 39 (84.8) | 0.29 |
| Interpersonal reason | 14 (26.7) | 31 (68.9) | 4 (12.9) | 27 (87.1) | 98 (17.0) | 477 (83.0) | |||
| More than one reason | 5 (41.7) | 7 (58.3) | 8 (16.0) | 42 (84.0) | 17 (24.6) | 52 (75.4) | |||
Row percentages add up to 100.
PRIMER, Preventing Injecting by Modifying Existing Responses.
Table 3 presents findings from the multivariable logistic regression models constructed for each study site among PWID who had provided previous injection drug initiation assistance in their lifetime. In Vancouver, those who previously assisted in the initiation of a family member or intimate partner were less likely to report self-perception of assisting with future injection drug use initiation compared to those who assisted friends [adjusted odds ratio (AOR) 0.4; confidence intervals (CI) 0.2, 0.8]. Additionally, daily injection drug use (AOR 3.6; 95% CI 2.1, 6.4) and less than daily injection drug use (AOR 2.6; 95% CI 1.4, 2.9) was associated with self-perception of assisting with future injection drug use initiation compared to those who did not inject in the past 6 months.
Table 3.
Multivariable logistic regression to assess the association between relationship to past initiates and self-perception of assisting with future injection drug use initiation among people who inject drugs who have previously assisted in initiation in their lifetime in Tijuana, San Diego and Vancouver (PRIMER; n = 933)
| Tijuana, Mexico (n =72) | San Diego, USA (n =85) | Vancouver, Canada (n=693) | |||||||
|---|---|---|---|---|---|---|---|---|---|
| AOR (95% CI) | X2 | P | AOR (95% CI) | X2 | P | AOR (95% CI) | X2 | P | |
| Age, years | 1.02 (0.96−1.08) | 0.53 | 0.47 | 1.02 (0.97−1.07) | 0.80 | 0.37 | 0.98 (0.97−1.0) | 2.75 | 0.10 |
| Sexa | |||||||||
| Female | Ref | - | - | Ref | - | - | Ref | - | - |
| Male | 2.06 (0.55−7.78) | 1.14 | 0.29 | 2 (0.48−8.33) | 0.90 | 0.34 | 1.81 (1.14−2.85) | 6.43 | <0.05 |
| Drug injection frequencyb | |||||||||
| None | - | - | - | Ref | - | - | Ref | - | - |
| Less than daily | - | - | - | 5.90 (1.29−27.10) | - | - | 2.56 (1.42−2.85) | - | - |
| Daily | 2.07 (0.37−11.35) | 5.71 | 0.06 | 3.63 (2.05−6.44) | 19.50 | <0.001 | |||
| Relationship to past initiates | |||||||||
| Friends | Ref | - | - | Ref | - | - | Ref | - | - |
| Family member/intimate partner | - | - | - | 0.11 (0.01−2.55) | - | - | 0.37 (0.16−0.84) | - | - |
| Other | 12.0 (2.06−70.32) | - | - | 0.29 (0.01−8.18) | - | - | 1.38 (0.66−2.88) | - | - |
| More than one relationship | 1.04 (0.18−6.04) | 7.70 | <0.05 | 0.90 (0.24−3.34) | 2.33 | 0.51 | 1.14 (0.64−2.03) | 7.26 | 0.06 |
In San Diego, three participants identified as transgender and were grouped into the female category.
Past six months; injection frequency was not included in the Tijuana model because most of the subjects reported daily use (>90%).
AOR, adjusted odds ratio; CI, confidence interval; PRIMER, Preventing Injecting by Modifying Existing Responses.
In San Diego, there was no significant association with the relationships with the initiate and self-perception that they will assist with future injection drug use initiation. Reporting injecting less than daily in the past 6 months was associated with higher odds of self-perception that they will assist with future injection drug use initiation compared to those who did not report any injection drug use (AOR 5.9; 95% CI 1.3, 27.1).
In Tijuana, those who provided injection initiation assistance to someone that fell into the category of ‘other’ had higher odds of self-perception that they will assist with injection drug use initiation compared to those who assisted friends (AOR 12.0; 95% CI 2.1, 70.3).
DISCUSSION
In a multi-site analysis undertaken across cities in three countries of North America, we investigated participants’ self-perception that they will assist with future injection drug use initiation. While findings in San Diego, Tijuana and Vancouver varied, both a person’s relationship to those they have provided initiation assistance to in the past and injection frequency emerged as key risk factors. These should therefore be considered when developing prevention interventions to attenuate the risk of injection drug use initiation. To our knowledge, this is the first paper to quantitatively model PWID’s self-perception of assisting with future injection drug use initiation.
As with all observational and multi-site studies, particularly those that include independent cohort studies from multiple countries and locations, this study includes several limitations. First, study participants were not randomly selected; PWID are generally a hidden population and cannot be sampled using probability approaches. Due to the sampling methods we used, we cannot determine if the results are generalisable to other PWID within these sites. Additionally, there are important differences in the policy, spatial and socioeconomic environments by PWID in San Diego, Tijuana and Vancouver; this may have led to heterogeneity among participant experiences. Second, this study was limited to data from a single visit (i.e. the predictor and outcomes are from the same visit). As such, we are limited in our capacity to identify causal relationships and instead report on associations between relevant variables captured at a single moment in time. Third, the small sample size from Tijuana and San Diego limited our power to detect significant associations and in the case of Tijuana, resulted in a few inflated confidence intervals, suggestive of a high level of uncertainty in the reported associations. The small sample size also impacted our ability to expand our independent variables to encompass wider response categories. Finally, because we used self-report data, the number of participants who likely provided initiation assistance may be higher as a result of underreporting. Research indicates that social response bias and internalised stigma may impact the willingness of participants to accurately report drug-related behaviours [11, 13, 37]; this is particularly the case for initiation assistance provision, which we have previously demonstrated is highly stigmatised [11, 38, 39].
Although no single variable was consistently found to be associated with self-perception of assisting with future injection drug use initiation across settings, we observed important associations by site. Most notably, in Vancouver, those who had previously assisted in the injection initiation of a family member or intimate partner were less likely to have a self-perception that they will assist with future injection drug use initiation again. It is possible that the closeness or proximity of this relationship (compared to other relationships) exposed participants assisting with initiation to witness poor outcomes of loved ones post-initiation, thereby reducing their willingness to provide assistance in the future. Qualitative research should be conducted for further exploration. Additionally, injection frequency was significantly associated with self-perception of assisting with future injection drug use initiation in San Diego and Vancouver. It is possible that the higher proportion of participants not injecting in San Diego and Vancouver may also be a contributing factor to participants in these settings reporting a lower likelihood of not assisting someone in the future. As we have previously reported, lower frequency of injection drug use or reporting no injection drug use is associated with a highly reduced risk of injection drug use initiation assistance provision in San Diego [40].
While only significant in Vancouver, having assisted with the initiation of an intimate partner or family member was associated with lower odds of self-perception of assisting with future injection drug use initiation. In Tijuana, those who previously assisted in the injection initiation of someone in the ‘other’ category (e.g. a stranger) had higher odds of self-perception of assisting with future injection drug use initiation. In line with the SEM, the proximity of certain people (e.g. family, intimate partners, friends), the specific geographic location and the context (e.g. political climate, laws and policies, access to health care, treatment and services) in which a person lives directly influences their capacity to make and carry out decisions and can highlight their risk and/or protective factors [24]. Focusing on the interconnectedness of an individual and their surroundings is particularly important when identifying culturally and individually appropriate intervention strategies [25]. Our findings suggest that the importance of close relationships (e.g. family, romantic partners) in influencing the risk of injection drug use initiation assistance provision matters and it likely varies by setting or cultural context, and that this may therefore be an important avenue of intervention to consider [17]. This is also in line with the social learning model [26], which highlights how one’s social network (i.e. interpersonal influences) can impact the perceptions, knowledge and interest in injection initiation. Additionally, the social learning model highlights how an individual’s intrapersonal convictions or experiences can encourage them to participate in injection drug use-related treatment and behavioural interventions. Both the SEM and the social learning model highlight the multiple contexts a person may experience when making decisions related to initiating someone into injection drug use, which was reflected in our findings.
While our sample size for who had previously provided injection drug use assistance was somewhat small (less than 100 people at two sites), those in Vancouver who reported they inject drugs daily had almost a four-fold increase in the odds of self-perception of assisting with future injection drug use initiation; those in San Diego who reported they have injected drugs in the last 6 months (less than daily) had a six-fold increase in odds of self-perception of assisting with future injection drug use initiation. There may be multiple interpretations for this association. For example, it is possible that people who inject or participate in any kind of injection drug use may be more likely to have severe substance use disorders; living with an active substance use disorder might make avoiding assisting others in initiation more difficult, especially if offered money or drugs to support their own drug use needs [41, 42]. Addressing the needs of those who are high-risk due to high injection frequency can be accomplished through enrolment in medication-assisted treatment for drug dependence (e.g. methadone, buprenorphine), which has been shown to be associated with lower odds of initiating others [40, 43, 44].
Structural interventions can focus on both physical and social environmental risk factors [45]. While participation in interventions and treatments that address multiple levels (e.g. intrapersonal, interpersonal, structural) are shown to improve health outcomes among PWID, interventions focused on social support (e.g. therapy focused on a couple or family unit, network or peer-based interventions) have been linked to fewer episodes of injecting drugs, decreasing injection risk behaviours, lower internalised stigma, improved health outcomes and increased utilisation of injection drug use related care, such as needle exchange programs [46–50]. For example, a systematic review focused on HIV prevention found that peer initiatives (e.g. outreach to peer PWID, training of peer PWID) reduced needle sharing by 63% (odds ratio 0.37; 95% CI 0.20, 0.67) [51]. Other successful individually-focused interventions include ‘Break the Cycle’, an in-person one-on-one intervention study that has been shown to reduce episodes of injection drug use initiation [52]. Its focus is on reducing the role PWID play in initiating injection naïve persons into injection through skill building and is designed based on social learning model approaches [26, 53].
In sum, in this study of PWID in cities of three North American countries, the relationship to past initiates and injection frequency emerged as characteristics that might increase the self-perception among PWID that they will assist others in injection drug use initiation in the future. In some sites, frequent drug injecting was associated with higher odds that participants reported a self-perception of assisting with future injection drug use initiation. Interventions such as medication-assisted treatment for substance use disorders that can reduce injection drug use frequency may therefore have the potential to enhance the resilience of PWID to avoid assisting others in the process of injection initiation [40, 43]. Further investigation into the social contexts experienced by PWID populations is needed to inform policy and interventions to prevent injection drug use initiation and its related harms.
Supplementary Material
Acknowledgements
The authors gratefully acknowledge the contributions to this research by the study participants and staff across sites.
This work was supported by the National Institute of Drug Abuse (NIDA) Avenir Award DP2DA040256-01 (PI: DW), NIDA R37DA019829 (PI: SAS), NIDA R01DA031074 (PI: RSG), NIDA RO1DA038965 (MPI: RB, Kral) and NIDA RO1DA045545 (MPI: RB, Corsi). DW was further supported by a New Investigator Award from the Canadian Institutes of Health Research, an Early Researcher Award from the Ontario Ministry of Research, Information and Science, and the St. Michael’s Hospital Foundation. CR was supported by a Postdoctoral Fellowship from the Canadian Institutes of Health Research. VDUS is supported by NIDA grant U01DA038886, and the ACCESS Study is supported by NIDA grant U01-DA021525. KH is supported by a Canadian Institutes of Health Research New Investigator Award (MSH-141971), a Michael Smith Foundation for Health Research Scholar Award, and the St. Paul’s Foundation. M-JM is supported by NIDA (U01-DA0251525), a Canadian Institutes of Health Research New Investigator Award and a Michael Smith Foundation for Health Research Scholar Award. He is the Canopy Growth professor of cannabis science at the University of British Columbia, a position established thanks to arms’ length gifts to the university from Canopy Growth, a licensed producer of cannabis, and the Government of British Columbia’s Ministry of Mental Health and Addictions.
Footnotes
Conflict of Interest
The authors report no conflict of interest. All National Institutes of Health/Wellcome-funded projects and other funding sources are listed below.
REFERENCES
- [1].Degenhardt L, Peacock A, Colledge S et al. Global prevalence of injecting drug use and sociodemographic characteristics and prevalence of HIV, HBV, and HCV in people who inject drugs: a multistage systematic review. Lancet Glob Health 2017;5:e1192–e207. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [2].Mathers BM, Degenhardt L, Phillips B et al. Global epidemiology of injecting drug use and HIV among people who inject drugs: a systematic review. Lancet 2008;372:1733–45. [DOI] [PubMed] [Google Scholar]
- [3].Uusküla A, Barnes DM, Raag M, Talu A, Tross S, Des Jarlais DC. Frequency and factors associated with providing injection initiation assistance in Tallinn, Estonia. Drug Alcohol Depend 2018;188:64–70. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [4].Vlahov D, Fuller CM, Ompad DC, Galea S, Des Jarlais DC. Updating the infection risk reduction hierarchy: preventing transition into injection. J Urban Health 2004;81:14–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [5].Werb D, Bluthenthal RN, Kolla G et al. Preventing injection drug use initiation: state of the evidence and opportunities for the future. J Urban Health 2018;95:91–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [6].Roy E, Haley N, Leclerc P, Cédras L, Blais L, Boivin JF. Drug injection among street youths in Montreal: predictors of initiation. J Urban Health 2003;80:92–105. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [7].Sherman SG, Smith L, Laney G, Strathdee SA. Social influences on the transition to injection drug use among young heroin sniffers: a qualitative analysis. Int J Drug Policy 2002;13:113–20. [Google Scholar]
- [8].Small W, Fast D, Krusi A, Wood E, Kerr T. Social influences upon injection initiation among street-involved youth in Vancouver, Canada: a qualitative study. Subst Abuse Treat Prev Policy 2009;4:8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [9].Crofts N, Louie R, Rosenthal D, Jolley D. The first hit: circumstances surrounding initiation into injecting. Addiction 1996;91:1187–96. [DOI] [PubMed] [Google Scholar]
- [10].Rotondi NK, Strike C, Kolla G et al. Transition to injection drug use: the role of initiators. AIDS Behav 2014;18:486–94. [DOI] [PubMed] [Google Scholar]
- [11].Guise A, Horyniak D, Melo J, McNeil R, Werb D. The experience of initiating injection drug use and its social context: a qualitative systematic review and thematic synthesis. Addiction 2017;112:2098–111. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [12].Neaigus A, Gyarmathy VA, Miller M, Frajzyngier VM, Friedman SR, Des Jarlais DC. Transitions to injecting drug use among noninjecting heroin users: social network influence and individual susceptibility. J Acquir Immune Defic Syndr 2006;41:493–503. [DOI] [PubMed] [Google Scholar]
- [13].Wenger LD, Lopez AM, Kral AH, Bluthenthal RN. Moral ambivalence and the decision to initiate others into injection drug use: a qualitative study in two California cities. Int J Drug Policy 2016;37:42–51. [DOI] [PubMed] [Google Scholar]
- [14].Bluthenthal R, Wenger L, Chu D et al. Factors associated with being asked to initiate someone into injection drug use. Drug Alcohol Depend 2015;149:252–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [15].Bryant J, Treloar C. Initiators: an examination of young injecting drug users who initiate others to injecting. AIDS Behav 2008;12:885–90. [DOI] [PubMed] [Google Scholar]
- [16].Young AM, Havens JR, Rudolph AE. Examining the social context of injection drug use: social proximity to persons who inject drugs versus geographic proximity to persons who inject drugs. Am J Epidemiol 2017;186:970–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [17].Simmons J, Rajan S, McMahon JM. Retrospective accounts of injection initiation in intimate partnerships. Int J Drug Policy 2012;23:303–11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [18].Barnes DM, Des Jarlais DC, Wolff M, Feelemyer J, Tross S. A qualitative study of persons who inject drugs but who have never helped others with first injections: how their views on helping contrast with the views of persons who have helped with first injections, and implications for interventions. Harm Reduct J 2018;15:43. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [19].Bluthenthal R, Wenger L, Thing J et al. Factors associated with time to injection initiation among drug injectors. Drug Alcohol Depend 2014; 140:e15–e6. [Google Scholar]
- [20].Navarro S, Kral AH, Strike CS, Simpson K, Wenger L, Bluthenthal RN. Factors associated with frequency of recent initiation of others into injection drug use among people who inject drugs in Los Angeles and San Francisco, CA, USA, 2016–17. Subst Use Misuse 2019;54:1715–24. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [21].Centers for Disease Control. The social-ecological model: a framework for prevention: National Center for Injury Prevention and Control, Division of Violence Prevention, 2019. Available at: https://www.cdc.gov/violenceprevention/overview/social-ecologicalmodel.html.
- [22].Bronfenbrenner U Toward an experimental ecology of human development. Am Psychol 1977;32:513–31. [Google Scholar]
- [23].Bates G, Tod D, Leavey C, McVeigh J. An evidence-based socio-ecological framework to understand men’s use of anabolic androgenic steroids and inform interventions in this area. Drugs Educ Prev Policy 2019;26:484–92. [Google Scholar]
- [24].Estrada JN Jr, Huerta AH, Hernandez E, Hernandez RA, Kim SW. Socio-ecological risk and protective factors for youth gang involvement. In: Shapiro H, ed. The Wiley handbook on violence in education: Forms, factors, and preventions. Hoboken, New Jersey: John Wiley & Sons, Inc.; 2018. [Google Scholar]
- [25].Wold B, Mittelmark MB. Health-promotion research over three decades: the social-ecological model and challenges in implementation of interventions. Scand J Public Health 2018;46:20–6. [DOI] [PubMed] [Google Scholar]
- [26].Strike C, Rotondi M, Kolla G et al. Interrupting the social processes linked with initiation of injection drug use: results from a pilot study. Drug Alcohol Depend 2014;137:48–54. [DOI] [PubMed] [Google Scholar]
- [27].Rhodes T, Bivol S, Scutelniciuc O, Hunt N, Bernays S, Busza J. Narrating the social relations of initiating injecting drug use: transitions in self and society. Int J Drug Policy 2011;22:445–54. [DOI] [PubMed] [Google Scholar]
- [28].Werb D, Garfein R, Kerr T et al. A socio-structural approach to preventing injection drug use initiation: rationale for the PRIMER study. Harm Reduct J 2016;13:25. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [29].Hamida AB, Rafful C, Jain S et al. Non-injection drug use and injection initiation assistance among people who inject drugs in Tijuana, Mexico. J Urban Health 2018;95:83–90. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [30].Robertson AM, Garfein RS, Wagner KD et al. Evaluating the impact of Mexico’s drug policy reforms on people who inject drugs in Tijuana, B.C., Mexico, and San Diego, CA, United States: a binational mixed methods research agenda. Harm Reduct J 2014;11:4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [31].Richardson LA, Kerr TH, Dobrer S et al. Socioeconomic marginalization and plasma HIV-1 RNA nondetectability among individuals who use illicit drugs in a Canadian setting. AIDS 2015;29:2487–95. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [32].Strathdee SA, Palepu A, Cornelisse PGA et al. Barriers to use of free antiretroviral therapy in injection drug users. JAMA 1998;280:547–9. [DOI] [PubMed] [Google Scholar]
- [33].Wood E, Stoltz J-A, Montaner JSG, Kerr T. Evaluating methamphetamine use and risks of injection initiation among street youth: the ARYS study. Harm Reduct J 2006;3:18. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [34].Kerr T, Marshall BDL, Miller C et al. Injection drug use among street-involved youth in a Canadian setting. BMC Public Health 2009;9:171. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [35].National Association of Health Data Organizations. Statistical approaches for small numbers: addressing reliability and disclosure risk, 2004.
- [36].Williams R Analyzing rare events with logistic regression University of Notre Dame, 2018. Available at: https://www3.nd.edu/$rwilliam/.
- [37].Guise A, Melo J, Mittal M et al. A fragmented code: the moral and structural context for providing assistance with injection drug use initiation in San Diego, USA. Int J Drug Policy 2018;55:51–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [38].Kulesza M, Watkins KE, Ober AJ, Osilla KC, Ewing B. Internalized stigma as an independent risk factor for substance use problems among primary care patients: rationale and preliminary support. Drug Alcohol Depend 2017;180:52–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [39].Islam MM, Topp L, Conigrave KM et al. The reliability of sensitive information provided by injecting drug users in a clinical setting: clinician-administered versus audio computer-assisted self-interviewing (ACASI). AIDS Care 2012;24:1496–503. [DOI] [PubMed] [Google Scholar]
- [40].Mittal ML, Vashishtha D, Sun S et al. History of medication-assisted treatment and its association with initiating others into injection drug use in San Diego, CA. Subst Abuse Treat Prev Policy 2017;12:42. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [41].Mitra S, Rachlis B, Krysowaty B et al. Potential use of supervised injection services among people who inject drugs in a remote and mid-size Canadian setting. BMC Public Health 2019;19:284. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [42].Linton SL, Celentano DD, Kirk GD, Mehta SH. The longitudinal association between homelessness, injection drug use, and injection-related risk behavior among persons with a history of injection drug use in Baltimore, MD. Drug Alcohol Depend 2013;132:457–65. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [43].Vashishtha D, Mittal M, Werb D. The North American opioid epidemic: current challenges and a call for treatment as prevention. Harm Reduct J 2017;14:7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [44].Mittal ML, Jain S, Sun S et al. Opioid agonist treatment and the process of injection drug use initiation. Drug Alcohol Depend 2019;197:354–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [45].Degenhardt L, Mathers B, Vickerman P, Rhodes T, Latkin C, Hickman M. Prevention of HIV infection for people who inject drugs: why individual, structural, and combination approaches are needed. Lancet 2010;376:285–301. [DOI] [PubMed] [Google Scholar]
- [46].Birtel MD, Wood L, Kempa NJ. Stigma and social support in substance abuse: implications for mental health and well-being. Psychiatry Res 2017;252:1–8. [DOI] [PubMed] [Google Scholar]
- [47].Rivera AV, DeCuir J, Crawford ND, Amesty S, Lewis CF. Internalized stigma and sterile syringe use among people who inject drugs in New York City, 2010–2012. Drug Alcohol Depend 2014;144:259–64. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [48].Robb J, Chou C, Johnson L, Liao H, Tan S. Mediating effects of social support and coping between perceived and internalized stigma for substance users. J Rehabil 2018;84:14–21. [Google Scholar]
- [49].Stanton MD, Shadish WR. Outcome, attrition, and family–couples treatment for drug abuse: a meta-analysis and review of the controlled, comparative studies. Psychol Bull 1997;122:170–91. [DOI] [PubMed] [Google Scholar]
- [50].Lyles CM, Kay LS, Crepaz N et al. Best-evidence interventions: findings from a systematic review of HIV behavioral interventions for US populations at high risk, 2000–2004. Am J Public Health 2007;97: 133–43. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [51].Medley A, Kennedy C, O’Reilly K, Sweat M. Effectiveness of peer education interventions for HIV prevention in developing countries: a systematic review and meta-analysis. AIDS Educ Prev 2009;21:181–206. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [52].National Institutes of Health. Break the cycle: prevention for reducing initiation into injection drug use. In: Medicine USNLo, Bethesda, Maryland: ClinicalTrials; 2018. Available from: https://clinicaltrials.gov/ct2/show/NCT03502525. [Google Scholar]
- [53].Uuskula A, Talu A, Barnes D et al. Implementing an updated “break the cycle” intervention to reduce initiating persons into injecting drug use in an eastern European and a US “opioid epidemic” setting. AIDS Behav 2019;23:2304–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.