UBC Faculty Research and Publications

A socio-structural approach to preventing injection drug use initiation : rationale for the PRIMER study Werb, Daniel; Garfein, Richard; Kerr, Thomas; Davidson, Peter; Roux, Perrine; Jauffret-Roustide, Marie; Auriacombe, Marc; Small, Will; Strathdee, Steffanie A. Sep 15, 2016

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


52383-Werb_D_et_al_Socio_structural.pdf [ 717.44kB ]
JSON: 52383-1.0340735.json
JSON-LD: 52383-1.0340735-ld.json
RDF/XML (Pretty): 52383-1.0340735-rdf.xml
RDF/JSON: 52383-1.0340735-rdf.json
Turtle: 52383-1.0340735-turtle.txt
N-Triples: 52383-1.0340735-rdf-ntriples.txt
Original Record: 52383-1.0340735-source.json
Full Text

Full Text

RESEARCH Open AccessA socio-structural approach to preventinginjection drug use initiation: rationale forthe PRIMER studyDaniel Werb1,6* , Richard Garfein1, Thomas Kerr2, Peter Davidson1, Perrine Roux3, Marie Jauffret-Roustide4,Marc Auriacombe5, Will Small2 and Steffanie A. Strathdee1AbstractBackground: Injection drug use remains a primary driver of HIV and HCV-related harms globally. However, there isa gap in efforts to prevent individuals from transitioning into injecting. People who inject drugs (PWID) play a keyrole in the transition of others into injecting, and while behavioral interventions have been developed to addressthis phenomenon, socio-structural approaches remain unexplored. To that end, we hypothesize that certaininterventions designed to reduce injecting-related risk behaviors may also reduce the risk that PWID exposeand introduce others into injecting. Identifying the preventive potential of existing interventions will informbroader efforts to prevent injecting and related harms.Methods: The Preventing Injecting by Modifying Existing Responses (PRIMER) study is a multi-country mixedmethods study with an aim to investigate whether specific interventions (e.g., opioid substitution therapy, supervisedinjection facilities, stable housing, incarceration environments) and related factors (e.g., public injecting and gender)influence the likelihood that PWID initiate others into injecting. This study will (1) investigate the PWID participation ininjection initiation; (2) identify factors influencing the risk that PWID expose others to or facilitate injection initiation; (3)describe drug scene roles that increase the risk of PWID facilitating injection initiation; and (4) evaluate the impact ofstructural, social, or biomedical interventions on the risk that PWID facilitate injection initiation. It does so by poolingobservational data from cohort studies of PWID in six cities: Vancouver, Canada; San Diego, USA; Tijuana, Mexico; Paris,Marseille, and Bordeaux, France.Results: Team members are conducting a prospective, multi-site study of PWID (n = 3050) in North America andFrance that includes quantitative and qualitative data collection through four separate cohort studies of PWID (SanDiego, STAHR II; Tijuana, El Cuete IV; Vancouver, V-DUS; Bordeaux, Marseille, Paris and Strasbourg, COSINUS).Conclusions: PRIMER is the largest study of injection initiation to date and the first to investigate structural approachesto preventing injection drug use initiation. Findings have the potential to inform the development and scale up of newand existing interventions to prevent transitions into injecting.Trial registration: Preventing Injecting by Modifying Existing Responses (PRIMER), NIDA DP2-DA040256-01.Keywords: HIV prevention, Injection initiation, Natural history of injecting, Street youth, People who inject drugs, Multi-site study* Correspondence: dwerb@ucsd.edu1Division of Global Public Health, Department of Medicine, University ofCalifornia San Diego, San Diego, USA6Division of Global Public Health, University of California School of Medicine,University of California San Diego, 9500 Gilman Drive, La Jolla, CA92093-0507, USAFull list of author information is available at the end of the article© 2016 The Author(s). Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, andreproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link tothe Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.Werb et al. Harm Reduction Journal  (2016) 13:25 DOI 10.1186/s12954-016-0114-1BackgroundInjection drug use remains a primary source of HIV- andhepatitis C (HCV)-related harms [1], with a recent esti-mate suggesting that 15.9 million people inject drugs glo-bally. Among this population, three million (19 %) arebelieved to be HIV-positive [1], while ten million peoplewho inject drugs (PWID) are estimated to be HCV-positive [2], with prevalence among PWID populationsbelieved to exceed 60 % in 37 countries [2]. The medianage of injection initiation globally is 19, indicating thatmore than half of PWID initiate injecting as adolescentsor young adults [3]. Of concern, data also suggest that thewindow of opportunity to prevent blood-borne virustransmission among new injectors is limited, given thatthe majority of incident HCV and hepatitis B virus casesoccur within 1 year of initiation [4], which may be due toelevated incidence of risk behaviors such as syringe shar-ing among recently initiated injectors [1]. The increasedaddictive potential of injecting also creates challenges inpreventing risky drug-related behaviors [5]. As such,injecting remains a primary driver of HIV and viral hepa-titis epidemics worldwide. Recent increases in prescriptionopioid misuse in North America may be contributing to aheightened risk of injection initiation [6], given that theuse of prescription opioids such as oxycodone has beenshown to be associated with a heightened risk of initiatinginjecting [7].Studies to date suggest that the majority of injection ini-tiation events are facilitated, either directly or indirectly,by PWID [8–19]. For example, among a sample of streetyouth who reported injecting drugs in Vancouver, Canada(n = 369), almost three quarters reported being initiatedby other injectors [10]. Among a cohort of PWID inTijuana, Mexico (n = 1052), only 11 % reported initiat-ing injection alone [20]. In France, among the PWIDenroled in a cross-sectional study in five cities (n = 1077;Lille, Strasbourg, Paris, Bordeaux, and Marseille), 83 % ofparticipants reported having been initiated by another per-son [21]. A qualitative study undertaken among a sampleof newly initiated PWID in New York City (n = 54) alsofound that decisions to begin injecting were influenced byprevious exposure to injecting and that initiates activelysought out PWID to assist with initiation [12]. Determin-ing the local factors driving such trends is critical in devel-oping adaptable and effective interventions to preventepidemics of injecting.In particular, preventing injection initiation requiresan understanding of individual roles within drug scenesthat may place PWID at higher risk of initiating others.For example, “hit doctors” (i.e., individuals paid withmoney or drugs to inject others) may be asked for injec-tion assistance by customers at the time they purchasetheir drugs. Hit doctors are often present in and aroundrisky injecting environments such as shooting galleriesand public injecting venues [22] and may be sources ofinjecting education [23], placing them at higher risk ofbeing approached for injection initiation assistance. Im-portantly, the context of initiation may also differ basedon the gender of both initiates and initiators, with fe-male initiates being more likely to be injected by oldermale PWID, who are often intimate partners [24]. Thesefindings suggest that PWID facilitate transitions toinjecting in multiple ways: by exposing injection-naïvedrug users to injecting [10], by providing injecting edu-cation [12], and by directly injecting initiates [15]. Whilea minority of PWID report facilitating injection initiationevents, those who do may initiate numerous individuals.Among a sample of Californian PWID (n = 605), for ex-ample, 35 % (n = 214) reported never initiating othersinto injecting, with a total of 3271 individuals initiatedinto injecting [15]. Similarly, while only 17 % (n = 55) ofa sample of Australian PWID (n = 324) reported initiat-ing others, they initiated 128 new injectors [16].Reducing the exposure of injection-naïve drug users toinjecting practices is therefore likely to reduce their riskof transitioning into injecting. Interestingly, data suggestthat neighborhoods with high levels of public injectingare sites of increased population mixing between PWIDand non-injectors [14]. For example, non-injecting streetyouth in Vancouver residing in the downtown eastside,which is characterized by high rates of public injecting,reported a cumulative incidence of initiation after 3 yearsthat was over twice that reported by non-injecting streetyouth residing elsewhere in the city [14]. Evidence alsosuggests that the deployment of intense policing ofstreet-level drug markets is likely to intensify risky injec-tion practices [25] and may also increase the exposure ofnon-injectors to PWID by dispersing drug marketsspatially [26] or by increasing the risk that PWID willinject in public or semipublic venues, a phenomenonthat has been widely observed across a range of set-tings [27–35]. The incarceration of PWID likely alsoincreases the risk that they may expose or initiatenon-injectors to injecting, given the potential forpopulation mixing in prison [17] (this is of particularinterest given recent moves towards de-incarcerationof PWID in California and Mexico [36]).Although there is strong evidence that PWID play akey role in the initiation of other individuals into injec-tion drug use [13], there are few interventions that pre-vent injection initiation by focusing on the role of PWIDin this practice [13]. To date, these have by and largeemployed behavioral approaches that seek to directlyalter the individual-level decision-making of PWID atrisk of initiating others into injecting. There is thereforea dearth of research and intervention development fo-cused on modifying the social and structural factors thatin turn influence PWID to initiate others.Werb et al. Harm Reduction Journal  (2016) 13:25 Page 2 of 10We hypothesize that the interventions that reduce thefrequency of public injecting may have the secondarybenefit of preventing injection initiation by limiting theexposure of non-injectors to injecting. For instance, opi-oid substitution therapies (OST) including methadoneand buprenorphine maintenance therapy, supervised in-jection facilities (SIF), and stable housing for PWID haveall been shown to reduce the risk of PWID injecting inpublic venues [37]. If public injecting increases the ex-posure of non-injectors to injecting, the provision ofsuch interventions may therefore reduce the risk thatPWID expose injection-naïve drug users to injecting,thereby decreasing the latter’s risk of injection initiation.Within settings characterized by population mixing be-tween PWID and non-injectors, there is often a “code”or ethic among established injectors discouraging theinitiation of others, but facilitating initiation may be dif-ficult to avoid for PWID who may be offered money ordrugs to facilitate initiation events [9]; this may be par-ticularly acute in the context of individuals with sub-stance use disorders who lack resources to purchasedrugs or have inadequate access to addiction treatmentssuch as OST.Given the gaps in knowledge regarding injection initi-ation, as well as the lack of effective interventions focusedon this practice, a 5-year multi-country study has been de-veloped to investigate structural, social, and biomedicalapproaches to preventing injection initiation and to iden-tify individual pathways and drug scene roles associatedwith facilitating initiation. Herein, we describe the meth-odology as well as the participant and site characteristics.MethodsPreventing Injecting by Modifying Existing Responses:PRIMERThe Preventing Injecting by Modifying Existing Re-sponses (PRIMER) is a US National Institute on DrugAbuse-funded study (DP2- DA040256-01; PI: Werb)seeking to determine whether interventions effective inreducing public injecting can reduce the initiation ofothers into injecting and to then employ these findingsto develop a socio-structural interventional approach topreventing injection drug use.Conceptual frameworkRhodes’ risk environment framework allows for the in-vestigation of the social and structural environment ex-perienced by PWID and an identification of how thismay limit the capacity of PWID to avoid drug-relatedharm [38, 39], including high risky injection practicessuch as exposing injection-naïve drug users to injectingbehaviors [40]. The risk environment framework also al-lows for an examination of how interactions between so-cial, structural, physical, and political/economic influencesoperating at micro, meso, and macro levels of an individ-ual’s environment may shape their local drug scene andbehavior. This is particularly useful in considering theintersection of individual-level behaviors—in this case, ex-posing or initiating others into injecting—with a range ofsocial, structural, and economic factors operating at mul-tiple levels. PRIMER will examine the interaction ofindividual-level factors such as OST enrolment, housingstatus, spatial proximity to non-injectors, income level,and level of drug dependence; meso-level factors such asthe presence of shooting galleries, public injecting settings,SIF implementation, OST programs, and drug use pat-terns in local PWID populations; and macro-level factorssuch as the type of drug policies and housing policies inplace, along with the availability of illicit drugs (Fig. 1).This approach is consistent with interventional ap-proaches that consider points of maximum effect andthe interaction of factors, at the structural level(macro), environmental level (meso), and individuallevel (micro) [41].Fig. 1 Conceptual framework to investigate injection initiation events facilitated by established injectorsWerb et al. Harm Reduction Journal  (2016) 13:25 Page 3 of 10ApproachThe PRIMER study has four major aims across twophases. Phase I consists of mixed methods data collec-tion and analysis. Phase II consists of the evaluation ofan interventional approach to prevent injection initiationinformed by the findings of phase I. Aim 1 is to deter-mine the prevalence and characteristics of PWID partici-pation in injecting initiation across multiple study sites.Aim 2 is to assess factors potentially influencing the riskthat PWID facilitate injecting initiation. Primary factorsof interest are as follows: enrolment in OST, SIF access,stable housing, a history of incarceration, public injec-tion, and gender. Aim 3 is to qualitatively investigate in-dividual pathways and roles that increase the risk ofPWID facilitating injecting initiation. This involves ex-ploring local social norms across study sites related toinjecting others among PWID through in-depth inter-viewing. We will specifically investigate gender withinthe context of injecting initiation events, with a focus onintimate partnerships. We will also explore the role ofinjecting environments (e.g., shooting galleries, jails/prisons, SIF, public vs. private settings) and drug sceneroles (e.g., hit doctors, drug dealers, sex workers) thatmay heighten the risk that PWID facilitate injecting ini-tiation. Finally, participants’ own initiation events will beexplored to determine how they influence subsequentdecisions to initiate others. Finally, aim 4 (carried out inphase II) is to test the impact of an existing structural orbiomedical intervention on the risk that PWID will facili-tate injecting initiation. Based on the findings generatedthrough the exploration of aims 1–3, we will evaluate theimplementation and/or scale-up of an interventional ap-proach addressing the factors that most greatly influencethe risk that PWID initiate others into injecting.In phase I, quantitative and qualitative data on in-volvement in injection initiation collected from PWIDaged 18 years and older in three longitudinal cohortstudies in North America and one longitudinal cohortstudy in France are being pooled and analyzed. Thesedata are being used to determine the prevalence andcharacteristics of PWID participation in injection initi-ation across multiple study sites, to assess a range of fac-tors at multiple levels that potentially influence the riskthat PWID facilitate injection initiation, and to investi-gate individual pathways and roles that increase the riskof PWID facilitating injection initiation. This consists ofboth quantitative and qualitative data collection to ex-plore local contexts and social norms related to PWID’sinitiation of others into injecting in each study site.Study sitesWhile we hypothesize that structural, social, and bio-medical interventions to reduce injection-driven HIVand HCV transmission risk might also reduce the riskthat PWID facilitate injection initiation, we anticipatethat the impact of these interventions will be mediatedby unique local characteristics in different settings andamong discrete populations of PWID. PRIMER there-fore includes longitudinal data from four cohort studiesof PWID aged 18 and older located in urban settings inthe USA (PI Dr. Richard Garfein, Study to Assess HepatitisC Risk (STAHR II); San Diego, CA; NIH R01-DA031074);Tijuana, Mexico (PI Dr. Steffanie Strathdee, Proyecto ElCuete IV (ECIV); Tijuana; NIH R37-DA019829); Vancouver,Canada (PI Dr. Thomas Kerr, V-DUS; Vancouver; NIH R01-DA011591); and France (PIs Pr. Marc Auriacombe and Drs.Marie Jauffret-Roustide and Perrine Roux, COhorte pourl’évaluation des facteurs Structurels et INdividuels de l’USagede drogues (COSINUS); Bordeaux, Marseille, Paris, andStrasbourg; INSERM/MILDECA C14-26).Quantitative data collectionIdentical questions on the involvement of PWID in in-jection initiation were seeded into cohort survey instru-ments in September 2014 and have been integrated intobroader data collection protocols. Questions solicit dataon participants’ experiences with initiating others intoinjecting, the reasons for doing so, the relationship be-tween initiates and initiators (including their gender),and the self-perceived likelihood that participants wouldinitiate others in the future. These questions form thebasis of the project’s quantitative data collection and willallow for investigations of the lifetime history and recentexperiences (i.e., past 6 months) of PWID’s facilitation ofinjection initiation. Importantly, these cohort studies allseek to investigate HIV risk behaviors among PWID res-iding in urban centers, and their survey instrumentsshare important similarities that facilitate cross-sitecomparisons. Additionally, the STAHR II and El CueteIV study instruments were intentionally designed forcross-study comparisons [42]. Prior to data pooling, eachparticipant’s data is assigned a unique identifier (includinga site locator) to avoid duplication or misclassification.Recruitment for all participant cohort studies is viatargeted sampling methods that include outreach activ-ities, snowball sampling methods, and self-referral. Po-tential participants are invited to participate in thestudy, and in the case of STAHR II, El Cuete IV, andVIDUS II, those individuals who consent are asked toprovide a specimen for rapid HIV testing (COSINUSparticipants do not undergo serologic testing). Outreachstaff, including interviewers, frontline personnel, andfield staff, as well as nurses, are, in all cases, familiarwith local neighborhoods and the social structural con-texts of injection drug use and have extensive experiencein recruiting and following up with participants fromlocal PWID networks. The field offices employed forparticipant interviews and interactions are designed asWerb et al. Harm Reduction Journal  (2016) 13:25 Page 4 of 10low-threshold, easily accessible, and friendly environ-ments, with staff trained to be courteous and respectfulof participants. In all studies, staff involved—and par-ticularly research coordinators and outreach staff—haveextensive experience working with addicted and margin-alized individuals. Study interviewers are trained in sur-vey administration, techniques to ensure trust andrapport with participants, and methods to ensure confi-dentiality. All interviews take approximately 60–90 minand are conducted in secure field offices and localesPWID are known to frequent. All follow-up interviewsoccur at 6-month intervals.Statistical analytic planOur primary analytic approach employs generalized linearmixed models (GLMM), which addresses problems of at-trition, missing data, variable timing of subject visits, andother unintended violations of design. Temporal correla-tions between subject responses are assumed using anautoregressive structure, with subject-specific variancesand correlations assumed to decline exponentially overtime. Importantly, this approach also allows for the inclu-sion of study site as a nesting factor. To evaluate the im-pact of socio-structural factors on the risk that PWIDfacilitate injection initiation, GLMM models will be fittedwith reporting recent (i.e., past 6 months) injection initi-ation facilitation as the outcome. This is defined using thefollowing survey item: “In the last 6 months, have youhelped anybody inject who had never injected before?”(yes vs. no). The primary independent variables of interestare defined as follows: recent (i.e., past 6 months) OST en-rolment, recent attendance at a SIF (Vancouver only), re-cent housing stability, recent incarceration, recent publicinjecting, and gender. To investigate how local settingsimpact risk, the effect of the independent variables ofinterest on facilitating injection initiation is being evalu-ated separately for each location. We anticipate a range ofconfounders which may influence the association betweenour outcome variable and independent variables ofinterest. While these are likely to vary across sites, weanticipate testing the impact of the following: intimatepartnerships, migration, stigma, HIV status, mentalhealth, health-care utilization, level of drug depend-ence, income level, ethnicity, and attitudes regardingsocial norms around injecting.Qualitative data collectionIn-depth qualitative interviews will be conducted amongparticipants from all cohort studies. Sampling is purpos-ive and based on either reporting facilitated injection ini-tiation in quantitative interviews or with reporting noinjection initiation but having a quantitative “profile”(i.e., identical answers to a set of relevant cohort surveyitems) consistent with other participants that reportfacilitating injection initiation. This approach allows forqualitative comparisons between individuals who do anddo not initiate others into injecting despite a profile thatsuggests that they are at risk; this will facilitate the de-velopment of targeted interventional approaches. Wewill weight our sampling based on the experience withthe independent variables of interest. For example, toensure recruitment of hit doctors and drug dealers, wewill recruit participants who indicate in quantitative sur-veys that they have initiated others into injecting andhave also undertaken these drug scene roles. We willseek to recruit 25–35 participants from each cohort. Inqualitative interviews, we will specifically investigategendered dynamics within the context of injection initi-ation events, with a focus on intimate partnerships. Wewill also explore the role of injecting environments (e.g.,shooting galleries, jails/prisons, SIF, public vs. privatesettings) and drug scene roles (e.g., hit doctors, drugdealers, sex workers) that may heighten the risk thatPWID facilitate injection initiation. Finally, we will ex-plore the participants’ accounts of their own initiationevents to determine how these influence subsequent de-cisions to initiate others.Qualitative analysisGrounded theory will be used as the overarching con-ceptual approach to qualitative data collection and ana-lysis [43, 44]. Grounded theory is founded on anapproach that assumes that one’s communication andactions express a meaning that is both individually heldand shared socially [43]. As such, grounded theory al-lows for the identification and delineation of core pro-cesses or assumptions that underlie specific individualactions or social interactions, with the ultimate aim ofidentifying a coherent model for the individual and so-cial actions observed. In the context of identifying indi-vidual pathways toward injection initiation, as well asthe social relationships that may propel an individual tobegin injecting, grounded theory will be employed tocollect and analyze data in order to develop a model ofinjection initiation grounded in social interaction.Mixed methods analytic planA parallel mixed methods design is being employed,which is followed by a meta-inference process whereinresults of the quantitative and qualitative analyses arebeing synthesized [45]. This approach represents an“ethno-epidemiological” research design, as it integratestargeted qualitative research in relation to epidemio-logical cohort studies, and greater understanding of thelived experience of PWID can add a deeper social di-mension to traditional quantitative measures regardingspecific outcomes (in this case, injection initiation), ashas been done previously [46–48]. Because effect sizesWerb et al. Harm Reduction Journal  (2016) 13:25 Page 5 of 10for each independent variable of interest will differacross study sites, qualitative findings are being used todetermine which causal pathways, individual roles, or re-lationships shape these differences. The meta-inferenceprocess is iterative and, as data are collected and trian-gulated, will refine our understanding of the factors driv-ing PWID to facilitate initiation.Phase II intervention development strategyThe type of intervention, or combination, to be tested inphase II will be determined after the analysis of findingsfrom phase I. This tailored approach extends to the selec-tion of the most appropriate setting for evaluation. As out-lined above, important differences exist across study sites,with interventions being scaled up or implemented duringthe study period; the situation “on the ground” in eachstudy site, in combination with phase I findings, willtherefore determine which site is ultimately selected forevaluation and whether an existing, modified, or novelintervention is evaluated. While dependent on findings,we have identified methodological approaches appropriatefor intervention testing. Community randomization, withexperimental and delayed onset control arms, may be ap-propriate to test interventions such as subsidized OST orSIF. Interrupted time-series analyses may also be suitableto test widespread policy change (e.g., de-incarceration).Discrete HIV epidemics are driven by local dynamics,and we therefore anticipate that the findings generated inphase I will likely demonstrate that a combination ofsocio-structural factors heighten the risk that PWID facili-tate injection initiation, and thereby contribute to a higherpopulation-level risk of HIV transmission. As above, rele-vant potential factors vary greatly by study setting andhave implications for the impact and optimal combinationof interventions. For example, in Vancouver, a combin-ation of OST and SIF access may reduce the risk thatPWID initiate others; in Tijuana, access to stable housingand de-incarceration may be found to most greatly reducerisk. In San Diego, with low OST enrolment and highlevels of homelessness, and a rapid move towards de-incarceration [49], multiple potential interventional pointsexist. In France, OST coverage is very high (70–80 %) (andhas been accompanied by a reduction in the overall inci-dence of sharing behaviors and HIV and HCV transmis-sion in local contexts [50, 51]); this provides anopportunity to determine how scale up of OST may im-pact injection initiation risk.Study sitesTable 1 provides relevant study site and baseline partici-pant characteristics for each cohort study included inPRIMER. These data do not include COSINUS sampledata given that data collection for this cohort has not yetstarted. We also present data from the VIDUS II cohort,which is being transitioned into the V-DUS study in2016 in order to integrate cohorts of PWID and street-involved youth, some of whom are injection-naïve. Theestimated sample sizes for each cohort are as follows:STAHR II, n = 575; Proyecto El Cuete, n = 750; V-DUS,n = 1735; and COSINUS, n = 680. As such, the pooledsample size for PRIMER will approach 3740 participants,though we caution that missing data, loss to follow-up,and incompatibility between variables in cohort ques-tionnaires are likely to reduce this overall number. Ascan be seen in Table 1, female participants comprise be-tween 26 and 38 % of the cohort participants, and HIVprevalence at baseline ranged from 0–10 % (note thatonly HIV-seronegative PWID participants are eligible forinclusion in VIDUS II, though V-DUS will not includethis restriction). Overall, the prevalence of lifetime his-tory of accessing OST was highest in Vancouver (53 %)and lowest in Tijuana (28 %). In Vancouver, 23 % of par-ticipants also reported accessing OST in the past6 months. We anticipate that reported OST access willbe higher among COSINUS participants given muchhigher levels of OST scale up in France [52].Ethics, consent, and permissionsThe study was provided ethical approval by the Institu-tional Review Board of the University of California, SanTable 1 Baseline cohort characteristicsStudy Location Number Eligible agerangeStudy period OST uptake SIFpresentHomeless in thepast 6 months (%)Incarceration HIV+ insample (%)Female(%)Proyecto ElCuete IVTijuana, Mexico 750 18+ 2010–2020 28 % ever No 20 71 % ever 5 38STAHR II San Diego, USA 575 18+ 2012–2016 23 % ever No 55 78 % ever 10 26V-DUSa Vancouver, Canada 1735 18+ 1996–2020 53 % ever Yes 5 19 % ever 0d 31COSINUSb Paris, Marseille,Bordeaux, andStrasbourg, France680 18+ 2016–2018 – Yesc – – – –aRefers to current VIDUS II sample (set to merge into V-DUS in 2016)bData collection to begin in Fall 2016cSIF likely to be implemented in Paris and Strasbourg in Fall 2016dAll VIDUS II participants are HIV-negative at baselineWerb et al. Harm Reduction Journal  (2016) 13:25 Page 6 of 10Diego School of Medicine (UCSD IRB 150866). We havealso obtained consent to publish and report individualparticipant data.Results and discussionInjection drug use remains one of the key drivers of HIVand HCV epidemics worldwide. And yet, very few inter-ventions have proven effective in preventing individualsfrom initiating injection drug use. Although experts sup-port efforts to reduce injecting exposure throughindividual-level behavioral change among PWID [13],structural approaches to reducing the risk that PWIDexpose and initiate others into injecting have not yetbeen adequately developed. While preliminary datastrongly suggest that PWID play a central role in facili-tating injection initiation, a critical gap in programing toaddress this practice remains.PRIMER, which seeks to harness the preventive poten-tial of existing structural, social, and biomedical HIV in-terventions, is, to our knowledge, unique in addressingthis gap in public health programing. This is highly rele-vant given the potential for increases in injection druguse presented by the rising prevalence of prescriptionopioid misuse in North America [6], the emergence andexpansion of epidemics of injecting in West and Sub-Saharan Africa and Asia [53], and the potential impactof widespread OST coverage on mitigating injection ini-tiation risk. We also anticipate that the mixed methodsanalytic approach we employ will further confirm genderand drug market roles (e.g., hit doctors) as uniquely con-tributing to the risk of injection initiation posed byPWID [11, 12]; this may allow for the development ofgender-specific interventions to prevent injection initi-ation. A peer-based intervention tailored to other spe-cific subpopulations may be required to reduce the riskthat certain PWID initiate others; these might includeBreak the Cycle or the “heroin sniffers” projects (bothbehavioral interventions that seek to build resilienceamong PWID to avoid initiating others) [13], which arecurrently being studied in combination in a research trialbeing carried out in New York City, USA, and Talinn,Estonia [54]. Relatedly, the potential that hit doctorsmay be more likely to initiate others might also requireadaptation of existing injection education interventions[55]. These models generally involve peer educators whodisseminate information about safer injection practices[56]; as such, they could be used to also disseminatestrategies to discourage injection initiation, as well as toprovide information to PWID on accessing existing in-terventions such as OST and SIF.With respect to the selection of study sites, the deci-sion was based primarily on a recognition, as above, thatpatterns of injection drug use initiation are driven byunique local dynamics. We therefore sought to includesites that represented diverse economic, legal, policy, andcultural settings. We note that while these site characteris-tics can be identified a priori, other important factors thatmay influence patterns of injecting initiation—particularlystigma, drug use trends, discrete social networks, and so-cial norms around injecting—will be identified during thestudy and will be used to inform intervention develop-ment. Specifically, Vancouver has a large, densely locatedPWID population, with a high but declining prevalence ofHIV [57]. In the mid-1990s, drug use patterns amongPWID shifted from heroin to cocaine injection, which wasundertaken with higher frequency, resulting in higherrates of contaminated syringe sharing and a subsequentincrease in HIV incidence [58]. Subsequently, a range ofharm reduction and ancillary services were implementedor scaled-up, such as OST and a limited number of SIFs[57], including a peer-run unsanctioned SIF that allowsassisted injection [59]. The existing of a peer-run SIF mayallow for intervention among individuals such as hit doc-tors who may more frequently engage in injection initi-ation events [60].San Diego is characterized by a PWID population esti-mated most recently at approximately 20,000 (standarddeviation 12,000) [61] highly dispersed across thecounty. PWID in San Diego report primarily injectingwith heroin and crystal methamphetamine [62]. HIVprevalence among PWID is estimated at 4 % [42], andaccess to services such as needle exchange and OST islow. Importantly, California has had among the moststringent criminal justice responses to drug use and pos-session since 1994 and has experienced extreme prisonovercrowding [49]. However, the passage in November2014 of Proposition 47, a statute that reclassifies non-violent drug crimes from felonies to misdemeanors, willresult in the release of potentially 10,000 currently incar-cerated individuals and is estimated to contribute to40,000 fewer felony convictions annually [49]. This sitetherefore presents an opportunity to assess the impact ofa large-scale structural intervention on injection initi-ation risk, in a setting currently characterized by high in-carceration rates.Tijuana has a densely situated local PWID population(estimated N = 10,000), located primarily within theZona Norte neighborhood (adjacent to the US-Mexicoborder) [63], with a high level of population mixing be-tween PWID and non-injectors as a result of deportationfrom the USA [64, 65]. HIV prevalence is low (i.e., 4 %)[42], and addiction treatment and harm reduction ser-vices are under-resourced [66]. However, the passage ofa partial drug decriminalization law in 2009 has seen agradual shift towards increased addiction treatmentscale-up (including OST) in Tijuana [67]. Accordingly,the number of OST clinics has increased slightly, whileshifting policing strategies are beginning to impact theWerb et al. Harm Reduction Journal  (2016) 13:25 Page 7 of 10rate of diversion of PWID towards addiction treatmentservices [36]. The STAHR II and El Cuete IV study in-struments were deliberately designed to be identical toallow cross-border comparisons of PWID in these adja-cent cities [42].France (Bordeaux, Marseille, Paris, and Strasbourg) isunique in having high levels of OST scale-up, with an es-timated 180,000 individuals reporting OST access [50](in specific sites, estimates of OST access among PWIDexceed 70 %) [68, 69]. Further, despite delays, SIF are ex-pected to be implemented in at least one study site (i.e.,Paris) during the project period [70, 71], providing anopportunity to investigate how a combination of highOST coverage and SIF access may impact PWID risk offacilitating initiation.Strengths and limitationsThe employment of existing NIDA-funded cohort studymechanisms ensures that data are of high quality andthat sufficient statistical power for analyses is obtained.The focus on multiple factors (i.e., OST enrolment, SIFaccess, housing status, incarceration, public injecting,and gender), in isolation and in combination, also allowsfor a high degree of adaptation to local contexts. It alsoprovides a foundation for future modeling of the broaderimpact of these interventions on injection drug use pre-vention in a variety of settings.This project, though, has limitations. First, it is likelythat participation in injection initiation may be underre-ported by PWID given the stigma associated with thisbehavior [12], which may result in threats to statisticalpower or misclassification of data. However, the large es-timated sample size (n = 3760) provides a large amountof statistical power. Second, there may be discrepanciesin data collection across sites that may limit our analyses(i.e., ECIV and STAHR II both employ computer-assisted personal interviewing (CAPI), while V-DUS doesnot (nor will COSINUS), which may result in misclassifi-cation given that recall of risk behaviors may differ be-tween cohort participants interviewed via CAPI andthose interviewed using other techniques) [72–74].However, as noted, all study questionnaires are highlycomparable, and we are undertaking extensive qualita-tive data collection as part of a mixed methods approachto obtain a more contextualized and detailed under-standing of the roles, social influences, and contexts thatshape PWID risk of facilitating injection initiation.Third, while study samples may not be generalizable,data from multiple settings allows for an understandingof how socio-structural factors in combination impactthe risk of injection initiation and, ultimately, of HIVtransmission. Fourth, challenges with scale-up exist withall interventions. However, team members have exten-sive experience in implementing and evaluating cutting-edge HIV interventions [75] and have also developed in-stitutional and multi-sectoral partnerships to supportsuch activities in all study sites.ConclusionsThere is a critical gap in the global development andscale-up of evidence-based responses to preventing injec-tion initiation. Given the contribution of injecting to therisk of transmission of HIV and other blood-borne dis-eases, it is imperative that resources be allocated towardsidentifying effective approaches to preventing injectiondrug use, rather than focusing solely on injection-relatedhealth harms. To ensure maximum effectiveness, any pre-ventive approach must also be both adaptable to a rangeof settings and easily integrated into the suite of currentpublic health programing designed to address the needs ofdrug-using populations [76]. These are the collective goalsof PRIMER, and we anticipate that study findings will helpsupport the development and evaluation of a comprehen-sive and scalable approach to reducing the incidence ofinjection-related morbidity and mortality experienced in awide range of settings internationally [1, 77–79].AbbreviationsCAPI: Computer-assisted personal interviewing; COSINUS: COhorte pourl’évaluation des facteurs Structurels et INdividuels de l’USage de drogues;ECIV: Proyecto El Cuete IV; HCV: Hepatitis C virus; HIV: Human immunodeficiencyvirus; OST: Opioid substitution therapy; PRIMER Study: Preventing Injecting byModifying Existing Responses Study; PWID: People who inject drugs;SIF: Supervised injection facility; STAHR II: Study to Assess Hepatitis Risk II;V-DUS: Vancouver Drug Users StudyAcknowledgmentsNot applicable.FundingDan Werb is supported by a US National Institute on Drug Abuse AvenirAward (DP2-DA040256-01) and a Canadian Institutes of Health Research NewInvestigator Award. Support for this project also comes from NIDA grantsR37-DA019829, R01-DA031074, R01-DA011591, and U01-DA03888602 and viaMILDECA support for COSINUS (INSERM/MILDECA C14-26).Availability of data and materialsData and materials cannot be shared in accordance with the need to ensurethe confidentiality of study participants.Authors’ contributionsDW conceived and designed the study, with methodological support fromSS, PD, WS, and RG. RG, TK, SS, PR, MA, and MJR provided the participantdata. DW drafted the manuscript, and all authors provided substantialrevisions. All authors read and approved the final manuscript.Competing interestsThe authors declare that they have no competing interests.Authors’ informationNot applicable.Consent for publicationNot applicable.Ethics approvalThe study was provided ethical approval by the Institutional ReviewBoard of the University of California, San Diego School of Medicine(UCSD IRB 150866).Werb et al. Harm Reduction Journal  (2016) 13:25 Page 8 of 10Author details1Division of Global Public Health, Department of Medicine, University ofCalifornia San Diego, San Diego, USA. 2Urban Health Research Initiative,British Columbia Centre for Excellence in HIV/AIDS, Vancouver, Canada.3INSERM, UMR_S 912, Sciences Economiques & Sociales de la Santé etTraitement de l’Information Médicale (SESSTIM), F-13385 Marseille, France.4Inserm U988/CNRS UMR 8211, Ecole des Hautes Etudes en SciencesSociales, Université de Paris Descartes, Paris, France. 5SANSPY/CNRS USR3413, Université de Bordeaux, Bordeaux, France. 6Division of Global PublicHealth, University of California School of Medicine, University of CaliforniaSan Diego, 9500 Gilman Drive, La Jolla, CA 92093-0507, USA.Received: 3 May 2016 Accepted: 1 August 2016References1. Mathers BM, et al. Global epidemiology of injecting drug use and HIVamong people who inject drugs: a systematic review. Lancet. 2008;372(9651):1733–45.2. Nelson PK, et al. Global epidemiology of hepatitis B and hepatitis C inpeople who inject drugs: results of systematic reviews. Lancet. 2011;378(9791):571–83.3. WHO, HIV and young people who inject drugs: a technical brief. Geneva:Inter-Agency Working Group on Key Populations; 2014.4. Garfein RS, et al. Viral infections in short-term injection drug users: theprevalence of the hepatitis C, hepatitis B, human immunodeficiency, andhuman T-lymphotropic viruses. Am J Public Health. 1996;86(5):655.5. Volkow ND. Science of addiction: drugs, brains, and behavior. Collingdale:DIANE Publishing; 2009.6. Strathdee SA, Stockman JK. Epidemiology of HIV among injecting and non-injecting drug users: current trends and implications for interventions. CurrHIV/AIDS Rep. 2010;7(2):99–106.7. Young AM, Havens JR. Transition from first illicit drug use to first injectiondrug use among rural Appalachian drug users: a cross-sectional comparisonand retrospective survival analysis. Addiction. 2012;107(3):587–96.8. Crofts N, et al. The first hit: circumstances surrounding initiation intoinjecting. Addiction. 1996;91(8):1187–96.9. Small W, et al. Social influences upon injection initiation among street-involved youth in Vancouver, Canada: a qualitative study. Substance AbuseTreat Prev Pol. 2009;4(1):1.10. Werb D. Injection career trajectories among illicit drug users in Vancouver,Canada, in School of Population and Public Health. Vancouver: University ofBritish Columbia; 2013. p. 156.11. Goldsamt LA, et al. Circumstances, pedagogy and rationales for injectioninitiation among new drug injectors. J Commun Health. 2010;35(3):258–67.12. Harocopos A, et al. New injectors and the social context of injectioninitiation. Int J Drug Policy. 2009;20(4):317–23.13. Werb D, et al. Interventions to prevent the initiation of injection drug use: asystematic review. Drug Alcohol Depend. 2013;133(2):669–76.14. Chami G, et al. Neighborhood of residence and risk of initiation intoinjection drug use among street-involved youth in a Canadian setting. DrugAlcohol Depend. 2013;132(3):486–90.15. Kral, A., et al., Initiating people into illicit drug injection, in The College onProblems of Drug Dependence 76th Annual Meeting. 2014: Costa Rica.16. Bryant J, Treloar C. Initiators: an examination of young injecting drug userswho initiate others to injecting. AIDS Behav. 2008;12(6):885–90.17. Boys A, et al. Drug use and initiation in prison: results from a national prisonsurvey in England and Wales. Addiction. 2002;97(12):1551.18. Strike C, et al. Interrupting the social processes linked with initiation ofinjection drug use: results from a pilot study. Drug Alcohol Depend. 2014;137:48–54.19. Roy E, et al. Drug injection among street youth: the first time. Addiction.2001;97:1003–9.20. Morris MD, et al. “Injection first”: a unique group of injection drug users inTijuana, Mexico. Am J Addict. 2012;21(1):23–30.21. Jauffret-Roustide M, et al. A national cross-sectional study among drug-usersin France: epidemiology of HCV and highlight on practical and statisticalaspects of the design. BMC Infect Dis. 2009;9(1):113.22. McElrath K, Harris J. Peer injecting: implications for injecting order andblood-borne viruses among men and women who inject heroin. J SubstUse. 2013;18(1):31–45.23. Murphy S, Waldorf D. Kickin'down to the street doc: shooting galleries inthe San Francisco Bay Area. Contemp Drug Probs. 1991;18:9.24. Doherty MC, Edgar MPH, Monterroso MD. Gender differences in the initiationof injection drug use among young adults. J Urban Health. 2000;77(3):396–414.25. Werb D, et al. Effects of police confiscation of illicit drugs and syringesamong injection drug users in Vancouver. Int J Drug Pol. 2008;19(4):332.26. Werb D, Kerr T, Fast D, Qi J, Montaner JS, Wood E. Drug-related risks amongstreet youth in two neighborhoods in a Canadian setting. Health & Place.2010;16(5):1061–7.27. Rhodes T, et al. Situational factors influencing drug injecting, risk reductionand syringe exchange in Togliatti City, Russian Federation: a qualitativestudy of micro risk environment. Soc Sci Med. 2003;57(1):39.28. Reuter P. The unintended consequences of drug policies, in RAND technicalreport. Santa Monica: RAND Europe; 2009. p. 36.29. Drucker E. Drug prohibition and public health: 25 years of evidence. PublicHealth Rep. 1999;114(1):14.30. Wood E, et al. Vienna Declaration: a call for evidence-based drug policies.Lancet. 2010;6736(10):2.31. Kori N, et al. Correlates of injecting in an HIV incidence hotspot amongsubstance users in Tijuana, Mexico. Int J Drug Policy. 2013;25(3):8.32. Kerr T, Small W, Wood E. The public health and social impacts of drug marketenforcement: a review of the evidence. Int J Drug Policy. 2005;16(4):210.33. Small W, et al. Public injection settings in Vancouver: physical environment,social context and risk. Int J Drug Policy. 2007;18(1):27–36.34. Wood E, et al. Displacement of Canada's largest public illicit drug market inresponse to a police crackdown. CMAJ. 2004;170(10):1551.35. Maher L, Dixon D. Policing and public health: law enforcement and harmminimization in a street-level drug market. Brit J Criminol. 1999;39(4):488.36. Consejo Nacional contras las Adicciones, Normatividad y Legislación.Mexico City: CONADIC; 2010.37. Rhodes T, et al. Public injecting and the need for "safer environment interventions"in the reduction of drug-related harm. Addiction. 2006;101(10):1384.38. Rhodes T, et al. The social structural production of HIV risk among injectingdrug users. Soc Sci Med. 2005;61(5):1026.39. Rhodes T, et al. Drug injecting, rapid HIV spread, and the 'risk environment':implications for assessment and response. AIDS. 1999;13:S259–70.40. Rhodes T. The 'risk environment': a framework for understanding andreducing drug-related harm. Int J Drug Policy. 2002;13(2):85.41. Blankenship KM, et al. Structural interventions: concepts, challenges andopportunities for research. J Urban Health. 2006;83(1):59–72.42. Robertson AM, et al. Evaluating the impact of Mexico’s drug policy reforms onpeople who inject drugs in Tijuana, BC, Mexico, and San Diego, CA, UnitedStates: a binational mixed methods research agenda. Harm Reduct J. 2014;11(1):4.43. Byrne M. Grounded theory as a qualitative research methodology. Aorn J.2001;73(6):1155.44. Strauss A, Corbin JM. Basics of qualitative research: grounded theoryprocedures and techniques. Thousand Oaks: Sage Publications, Inc; 1990.45. Teddlie C, Tashakkori A. Foundations of mixed methods research:integrating quantitative and qualitative approaches in the social andbehavioral sciences. Thousand Oaks: Sage Publications Inc; 2009.46. McNeil R, et al. Hospitals as a ‘risk environment’: an ethno-epidemiologicalstudy of voluntary and involuntary discharge from hospital against medicaladvice among people who inject drugs. Soc Sci Med. 2014;105:59–66.47. Agar M. Toward a qualitative epidemiology. Qual Health Res. 2003;13(7):974–86.48. Lopez AM, et al. Interdisciplinary mixed methods research with structurallyvulnerable populations: case studies of injection drug users in SanFrancisco. Int J Drug Pol. 2013;24(2):101–9.49. The state of sentencing 2015: Developments in policy & practice. 2016: TheSentencing Project, Washington DC. Available from: http://sentencingproject.org/wp-content/uploads/2016/02/State-of-Sentencing-2015.pdf. Accessed June 2016.50. Fatseas M, Auriacombe M. Why buprenorphine is so successful in treatingopiate addiction in France. Curr Psychiatry Rep. 2007;9(5):358–64.51. Fatseas M, et al. Change in HIV-HCV risk-taking behavior and seroprevalenceamong opiate users seeking treatment over an 11-year period and harmreduction policy. AIDS Behav. 2012;16(7):2082–90.52. Mathers BM, et al. HIV prevention, treatment, and care services for peoplewho inject drugs: a systematic review of global, regional, and nationalcoverage. Lancet. 2010;375(9719):15.53. UNODC. World Drug Report 2015. Vienna: United Nations Office on Drugsand Crime; 2015.Werb et al. Harm Reduction Journal  (2016) 13:25 Page 9 of 1054. NIDA. 2015 Avant-Garde Awards offer extraordinary ideas in HIV/AIDSresearch. Bethesda: U.S. National Institute on Drug Abuse; 2015.55. Friedman SR, et al. Drug-scene roles and HIV risk among Puerto Ricaninjection drug users in East Harlem, New York and Bayamon, Puerto Rico. JPsychoactive Drugs. 2002;34(4):363.56. Small W, et al. The injection support team: a peer-driven program to addressunsafe injecting in a Canadian setting. Subst Use Misuse. 2012;47(5):491–501.57. Urban Health Research Initiative, Drug situation in Vancouver. Urban HealthResearch Initiative. Vancouver: BC Centre for Excellence in HIV/AIDS; 2013. p. 62.58. Tyndall MW, et al. Intensive injection cocaine use as a primary risk factor of HIVseroconversion among polydrug users in Vancouver. AIDS. 2002;17(6):887.59. McNeil R, et al. Negotiating place and gendered violence in Canada'slargest open drug scene. Int J Drug Pol. 2014;25(3):608–15.60. McNeil R, et al. “People knew they could come here to get help”: anethnographic study of assisted injection practices at a peer-run‘unsanctioned’ supervised drug consumption room in a Canadian setting.AIDS Behav. 2014;18(3):473–85.61. Friedman SR, et al. Estimating numbers of injecting drug users inmetropolitan areas for structural analyses of community vulnerability andfor assessing relative degrees of service provision for injecting drug users. JUrban Health. 2004;81(3):377–400.62. Roth AM, et al. Patterns of drug use, risky behavior, and health statusamong persons who inject drugs living in San Diego, California: a latentclass analysis. Subst Use Misuse. 2015;50(2):205–14.63. Strathdee SA, et al. The emerging HIV epidemic on the Mexico-US border:an international case study characterizing the role of epidemiology insurveillance and response. Annals Epidemiol. 2012;22(6):426–38.64. Brouwer KC, et al. Intraurban mobility and its potential impact on thespread of blood-borne infections among drug injectors in Tijuana, Mexico.Subst Use Misuse. 2012;47(3):244–53.65. Brouwer KC, et al. Spatial epidemiology of HIV among injection drug usersin Tijuana, Mexico. Ann Assoc Am Geogr. 2012;102(5):1190–9.66. Werb D, et al. Police bribery and access to methadone maintenancetherapy within the context of drug policy reform in Tijuana, Mexico. DrugAlc Depend. 2015;148:221–5.67. Beletsky L, et al. Implementing Mexico’s “Narcomenudeo” drug law reform:a mixed-methods assessment of early experiences among people whoinject drugs. J Mixed Methods Res, 2015. In Press.68. Roux P, et al. Retention in opioid substitution treatment: a major predictorof long-term virological success for HIV-infected injection drug usersreceiving antiretroviral treatment. Clin Infect Dis. 2009;49(9):1433–40.69. Weill-Barillet L, Pillonel J, Semaille C, Leon L, Le Strat Y, Pascal X, Barin F, Jauffret-Roustide M. Hepatitis C virus and HIV seroprevalences, sociodemographiccharacteristics, behaviors and access to syringes among drug users, a comparisonof geographical areas in France, ANRS-Coquelicot 2011 survey. Rev EpidemiolSante Publique. 2016. doi:10.1016/j.respe.2015.10.003.70. Arnaud D. Salle de shoot, le retard français, in Libération. Paris: Libération; 2014.71. Jauffret-Roustide M, Pedrono G, Beltzer N. Supervised consumption rooms:the French paradox. Int J Drug Pol. 2013;24(6):628–30.72. Ramjee G, Weber AE, Morar NS. Recording sexual behavior: comparison ofrecall questionnaires with a coital diary. Sex Transm Dis. 1999;26(7):374–80.73. Van Der Elst EM, et al. Is audio computer-assisted self-interview (ACASI)useful in risk behaviour assessment of female and male sex workers,Mombasa, Kenya? PLoS One. 2009;4(5):e5340.74. Des Jarlais DC, et al. Audio-computer interviewing to measure riskbehaviour for HIV among injecting drug users: a quasi-randomised trial.Lancet. 1999;353(9165):5.75. Garfein RS, et al. Methods to recruit and retain a cohort of young-adult injectiondrug users for the Third Collaborative Injection Drug Users Study/Drug UsersIntervention Trial (CIDUS III/DUIT). Drug Alcohol Depend. 2007;91:4–17.76. WHO. UNODC, UNAIDS technical guide for countries to set targets foruniversal access to HIV prevention, treatment and care for injecting drugusers: 2012 revision. Geneva: WHO/UNODC/UNAIDS; 2012.77. Degenhardt L, et al. The global epidemiology and burden of opioiddependence: results from the global burden of disease 2010 study.Addiction. 2014;109(8):1320–33.78. Mathers BM, et al. Mortality among people who inject drugs: a systematicreview and meta-analysis. Bull World Health Organ. 2013;91(2):102–23.79. Degenhardt L, et al. Mortality among regular or dependent users of heroinand other opioids: a systematic review and meta-analysis of cohort studies.Addiction. 2011;106(1):32–51.•  We accept pre-submission inquiries •  Our selector tool helps you to find the most relevant journal•  We provide round the clock customer support •  Convenient online submission•  Thorough peer review•  Inclusion in PubMed and all major indexing services •  Maximum visibility for your researchSubmit your manuscript atwww.biomedcentral.com/submitSubmit your next manuscript to BioMed Central and we will help you at every step:Werb et al. Harm Reduction Journal  (2016) 13:25 Page 10 of 10


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items