Antiretroviral therapy (ART) effectively suppresses viral load and dramatically reduces morbidity and mortality among persons living with HIV [1–3]. The apparent ability of ART to reduce HIV transmission among serodiscordant couples has also led to the increasingly well recognized notion that the expansion of access to ART may also be an effective HIV prevention strategy [4,5]. However, some have argued that expanded access to ART may result in increased HIV risk behaviour as a result of diminished concerns about HIV transmission among treated persons [6,7]. For example, some studies involving MSM  and injection drug users (IDUs)  have suggested that individuals who believe receiving ART protects against transmitting HIV are more likely to engage in unprotected intercourse and other HIV risk behaviours. Other studies among MSM have shown that initiating ART and immunologic response to treatment are associated with a return to high-risk sexual behaviours [10,11], which may contribute to HIV transmission .
Although international guidelines state that HIV-positive IDUs should have equitable and universal access to ART , the worldwide coverage of HIV prevention and treatment services for IDUs remains poor . In addition to the widely held view that IDUs are poor candidates for ART due to ongoing drug dependence and other co-morbidities , concerns have been raised that IDUs may resume high-risk behaviours following the initiation of HIV therapy . Previous studies investigating the relationship between ART initiation and increases in risk behaviour among IDUs are equivocal [16–18]. Although a European study showed significant decreases in risk behaviour following the commencement of HIV therapy , studies conducted in the United States have demonstrated marginal increases in sexual risk behaviour, particularly following successful immunologic response to therapy [16,17]. However, these studies were limited by relatively short follow-up periods (e.g., 2 years) and inconsistent adjustment for potential confounding variables (e.g., relationship status). Some studies have also relied on self-reported ART use, which has been shown to be a less-than-ideal measure of medication use among IDUs . Finally, studies conducted in settings where ART is not freely available may suffer from a selection bias due to financial barriers to accessing HIV care and treatment. We conducted this study using data from a long-running prospective cohort study of ART-naive HIV-positive IDUs to address the limitations of prior studies and to examine whether ART initiation was associated with increases in sexual risk behaviour in a setting where ART is freely available to eligible individuals.
Data for these analyses were derived from the AIDS Care Cohort to Evaluate Exposure to Survival Services (ACCESS), a community-recruited open prospective cohort of HIV-positive IDUs in Vancouver, Canada. The study methodology has been described in detail elsewhere [20,21]. In brief, recruitment began in May 1996 and occurred through extensive street-based outreach and word of mouth. Participants were eligible for the study if they were 18 years of age or older, resided in the greater Vancouver region, tested HIV-positive upon entry, had injected an illegal drug during the previous month, and provided informed consent. HIV infection was detected using ELISA and positive test results were confirmed using western blot. At baseline and at each semi-annual follow-up, participants completed a lengthy interviewer-administered questionnaire that elicits information regarding sociodemographic characteristics, drug use patterns, sexual behaviours (including all outcomes assessed in this study), and other relevant exposures. The present analysis is based on data for participants who completed a baseline and at least one follow-up interview between May 1996 and April 2008 and who were naive to ART upon entry into the cohort. This study was approved by the University of British Columbia/Providence Healthcare Research Ethics Board.
Clinical variables and exposure to antiretroviral therapy
Our setting is somewhat unique in that antiretroviral drugs have been centrally distributed at no cost to eligible individuals since 1986 . All eligible HIV-positive patients residing in British Columbia receive ART from the HIV/AIDS Drug Treatment Program, a province-wide dispensation programme and laboratory for HIV/AIDS clinical monitoring. ACCESS cohort data are linked in a confidential manner to the information collected by this programme. Thus, all CD4+ cell count and HIV RNA level measurements conducted throughout the study period are available from this database. CD4+ cell counts are determined by flow cytometry and plasma HIV RNA levels were measured using the Roche Amplicor Monitor assay (Roche Molecular Systems, Mississauga, Canada). Programme data also include a complete prospective profile of ART use, including the exact date of therapy initiation.
We examined the following sexual behaviours reported at baseline and at each subsequent interview. We defined ‘any sexual activity’ as engaging in voluntary sexual activity (i.e., oral, vaginal, or anal intercourse) with any same or opposite sex regular partner, casual partner, or sex trade client. Although sexual intercourse is not a risk factor for HIV transmission per se, sexually active participants represent a group at risk of engaging in high-risk behaviour. Participants were also asked to indicate whether they had engaged in any vaginal or anal intercourse without using a condom. Finally, we also examined the number of sex partners reported by participants and dichotomized this variable as greater than one (i.e., ‘multiple sex partners’) vs. less than or equal to one. All outcome variables refer to the 6 months prior to each semi-annual visit.
The objective of this study was to examine whether ART initiation was associated with increases in sexual activity or risk behaviours among HIV-positive IDUs. Therefore, the primary independent variable of interest was ‘initiate ART’, operationalized as an indicator variable representing the first follow-up interview between 6 and 12 months following the commencement of therapy. All other visits were coded as zero, as were all visits for individuals who never initiated therapy throughout the study period. The date of ART initiation was determined through a confidential database linkage as described above. To be conservative, follow-up interviews occurring within 6 months of initiating therapy were not considered an ART initiation visit, as behaviours assessed would have included periods of both very early ART treatment and no treatment.
We also examined variables that have been observed in other studies of HIV-positive IDUs to be associated with sexual behaviours [16–18]. Variables that were considered as potential confounders included age at baseline (per year older), sex (female vs. male), ethnicity (Aboriginal ancestry vs. other), current relationship status [partnered (i.e., married, common law, regular partner) vs. single or casually dating], homelessness (yes vs. no), sex trade work (yes vs. no), noninjection crack use (≥daily vs. <daily), injection heroin use (≥daily vs. <daily), injection cocaine use (≥daily vs. <daily), syringe sharing (yes vs. no), use of addiction treatment (yes vs. no), current enrolment in a methadone maintenance therapy (MMT) programme (yes vs. no), CD4+ cell count (per 100 cells/μl), and HIV-1 RNA plasma viral load (per log10 copies/ml). Unless otherwise indicated, all behaviours refer to the 6 months prior to the date of the interview. The year of interview was also included as a potential confounder, as we recognized that perceptions of ART, side-effects of treatment, and prevalence of risk behaviour may have changed over the study period. To further investigate possible changes in the impact of ART initiation over time, we tested for interaction effects between calendar year and the ‘initiate ART’ period.
We first examined whether those who initiated ART differed from those who did not initiate with respect to sociodemographic characteristics and baseline clinical factors. The Pearson's χ 2-test was used to test for differences between categorical variables and the Wilcoxon rank–sum test was used for continuous variables. We then performed a series of bivariate analyses using generalized linear mixed-effects modelling (GLMM). These longitudinal models allow for the analysis of both between-subject and within-subject variations in responses over time (i.e., individuals are assumed to have their own subject-specific trajectories) . Therefore, this modelling strategy permitted us to examine whether an individual's likelihood of engaging in sexual behaviour following ART initiation was significantly less or greater than his/her underlying propensity for engaging in these activities. To determine the independent association between initiating ART and sexual behaviour, we fit a series of multivariate models including all variables significant in bivariate analyses at an a priori defined cut-off of P less than 0.10. As has been suggested by other authors , we chose a more liberal criterion for the inclusion of covariates than the traditional cut-off of P less than 0.05 to ensure that all potentially confounding measured variables were included.
To further investigate the potential relationships between ART initiation, immunologic response to therapy, and return to engagement in sexual activity/risk behaviour, we conducted a series of sensitivity analyses. First, we hypothesized that individuals initiating ART with a CD4 cell count higher than 200 cells/μl would be more likely to engage in sexual behaviour than individuals initiating ART with a CD4 cell count below 200 cells/μl. To test this hypothesis, we created a categorical indicator variable consisting of four mutually exclusive levels: ‘initiate ART’ period with CD4 cell count 200 cells/μl or more, ‘initiate ART’ period with CD4 cell count less than 200 cells/μl, all other periods with CD4 cell count 200 cells/μl or more, and all other periods with CD4 cell count less than 200 cells/μl (referent). We calculated the bivariate associations between this variable and each sexual behaviour outcome. Second, to determine whether successful response to ART was associated with increases in sexual risk behaviour, we conducted a paired analysis among individuals with a CD4 cell count less than 200 cells/μl prior to ART initiation and a second measurement with a CD4 cell count 200 cells/μl or more after the commencement of therapy. We then compared self-reported sexual behaviour during these two periods using McNemar's exact test for matched data.
We also recognized that by examining an a priori defined ‘initiate ART’ period (i.e., between 6 and 12 months following the commencement of therapy), we may have failed to identify longer-term changes in sexual activity and risk. Therefore, we determined the proportion of participants reporting each outcome in the four follow-ups (i.e., 2 years) after the initiation of ART and examined changes over time using the Mantel trend test . All statistical analyses were conducted using SAS (version 9.1; SAS Institute, Cary, North Carolina, USA) and all P values are two-sided.
Between May 1996 and April 2008, 457 antiretroviral-naïve, HIV-positive participants were enrolled in the study and completed at least one follow-up interview. The median duration of follow-up was 52 months [interquartile range (IQR) 16 – 100]. Over the study period, 260 (56.9%) initiated ART. Of those, 197 (75.8%) completed an interview between 6 and 12 months following the date of ART initiation. The 63 participants who did not complete an interview between 6 and 12 months following the commencement of therapy and thus did not contribute behavioural data during the ‘initiate ART’ period did not differ with respect to age (P = 0.972), sex (P = 0.643), ethnicity (P = 0.101), or baseline involvement in any sexual activity (P = 0.615), unprotected sex (P = 0.960), or multiple sexual partnership (P = 0.878). The majority of individuals initiated ART prior to 1999 (median = 1998, IQR 1997–2002). Among those who initiated ART, the median number of follow-ups prior to and following initiation was 3 (IQR 1–6) and 7 (IQR 3–14), respectively. The median number of follow-ups among those who were never exposed to ART over the study period was 3 (IQR 2 – 8).
At baseline, the median age of the sample was 34.2 (IQR 27.7–40.8), 202 (44.2%) were women, and 178 (38.9%) were of Aboriginal ancestry. The sociodemographic characteristics of those who initiated ART over follow-up did not significantly differ compared with those who remained antiretroviral-naive (see Table 1). As expected, among ART initiates, the median baseline CD4 cell count was significantly lower (350 vs. 490 cells/μl, P < 0.001) and the baseline median HIV-1 RNA viral load was significantly higher (56 000 vs. 30 000 copies/ml, P < 0.001). At baseline, 331 (72.4%) reported sexual intercourse during the past 6 months, 158 (34.6%) reported recent unprotected vaginal or anal intercourse with a sex partner or client, and the median number of sex partners reported was 1 (IQR 1–2).
Any sexual activity
Among participants who completed an interview between 6 and 12 months after the initiation of ART, 101 (51.3%) reported having sexual activity. In a bivariate GLMM analysis, participants were no more likely to report sexual activity in the period following ART initiation than in other follow-ups [odds ratio (OR) = 1.02, 95% confidence interval (CI) 0.71–1.46, P = 0.921]. Other factors associated with sexual activity are shown in Table 2. In a multivariate model adjusting for sociodemographic characteristics and other variables, ART initiation remained nonsignificant [adjusted odds ratio (AOR) = 0.87, 95% CI 0.60–1.25, P = 0.453]. Female sex, partnered relationship status, daily cocaine use, and higher CD4 cell count were positively associated with sexual activity in the final multivariate model, whereas older age, current enrolment in MMT, and the year of interview were inversely associated with this outcome (see Table 2).
Within 6–12 months following ART initiation, 33 (16.8%) reported engaging in unprotected vaginal or anal intercourse with a client or sex partner. Table 3 shows the GLMM bivariate and multivariate analyses of engaging in unprotected intercourse over the follow-up period. The period following ART initiation was not associated with an increase or decrease in the likelihood of unprotected intercourse in either bivariate (OR = 0.79, 95% CI 0.49–1.26, P = 0.324) or multivariate (AOR = 0.82, 95% CI 0.51–1.31, P = 0.402) analyses. Factors significantly associated with unprotected intercourse in multivariate analysis were partnered relationship status, sex trade work, syringe sharing, and higher CD4 cell count (Table 3).
Multiple sex partners
Of those participants who initiated ART, 53 (26.9%) reported having multiple sex partners or clients during the 6–12-month period following the commencement of therapy. Even after extensive adjustment for potential confounders, ART initiation was not associated with having multiple sex partners (AOR = 0.93, 95% CI 0.61–1.40, P = 0.712). Female sex, daily crack use, daily cocaine use, and syringe sharing were positively associated with having multiple sex partners in multivariate analysis, whereas older age, partnered relationship status, current enrolment in MMT, and the year of interview were protective for this outcome (see Table 4). In all models, we observed no interaction between calendar year and ART initiation period (all P > 0.10).
To evaluate whether participants initiating ART with a CD4 cell count 200 cells/μl or higher were at an increased risk of engaging in sexual behaviour, we constructed a categorical indicator variable and examined the bivariate associations with each outcome. Relative to the non‘initiate ART’ periods in which CD4 cell counts were less than 200 cells/μl, sexual activity was more common during both the non‘initiate ART’ periods (OR = 1.94, 95% CI 1.56–2.41, P < 0.001) and ‘initiate ART’ period (OR = 1.76, 95% CI 1.13–2.76, P = 0.013) when CD4 cell counts were greater than 200 cells/μl. In contrast, unprotected sexual intercourse was only associated with the non‘initiate ART’, CD4 at least 200 cells/μl periods (OR = 1.50, 95% CI 1.13–1.98, P = 0.005) [i.e., ‘initiate ART’ period and having a CD4 measurement ≥200 cells/μl was not significant (P = 0.966)]. Similarly, having multiple sex partners was not associated with initiating ART and having a CD4 cell count 200 cells/μl or more (P = 0.752).
To further investigate the impact of immunological response to therapy on sexual behaviour, we conducted a subanalysis consisting of individuals with a CD4 measurement less than 200 cells/μl prior to initiation and a subsequent measurement 200 cells/μl or more after the commencement of therapy. Among these participants, 12 (19.1%) reported unprotected intercourse during the preinitiation period compared with 11 (17.5%) following ART initiation (McNemar's test P = 1.000). Multiple sex partners were reported by 17 (27.0%) and 10 participants (15.9%) during the preinitiation and postinitiation periods, respectively (McNemar's test P = 0.144).
As a final sensitivity analysis, we examined longer-term changes in sexual activity following ART initiation by calculating the proportion of individuals reporting each outcome during the four follow-up visits subsequent to the commencement of therapy. As shown in Fig. 1, there were no detectable trends in any sexual activity (P = 0.636), unprotected intercourse (P = 0.271), or multiple sexual partners (P = 0.380) over the 2-year period following ART initiation.
In this 12-year-long study of HIV-positive IDUs, ART initiation was not associated with increases in sexual activity or risk behaviour. Given these findings, concerns regarding potential increases in or resumption of sexual risk behaviour among IDUs following the initiation of HIV therapy appear to be unfounded in this setting.
Although previous research has shown that successful immunologic response to ART may lead to a resumption of engagement in sexual risk behaviour [10,16], we failed to observe this phenomenon in a series of subanalyses among participants initiating ART. Furthermore, although commencing therapy with CD4 cell count 200 cells/μl or more was associated with engagement in sexual activity, we did not detect an increased risk of unprotected intercourse or reporting multiple sexual partners. Longer-term sexual behaviour patterns following ART initiation were also stable in this cohort. Nonetheless, physicians that prescribe ART should be aware that although increases risk behaviour in the period directly following ART initiation may be unlikely, patients who exhibit improvements in health and social functioning should be provided with risk reduction education and counselling. Health professionals should also recognize the broader determinants of risk behaviour among IDU populations (including, e.g., the role of social networks, homelessness, incarceration, and other economic inequities [25–27]) and facilitate the engagement of patients initiating ART with other health and social services to address health disparities.
The present study has a number of limitations that should be noted. First, as this study was observational, access to ART was not random but influenced by a number of factors outside of our control, including, for example, physicians' beliefs regarding the prescription of ART to individuals with ongoing drug use . Although it is, therefore, possible that selection bias may have affected our results, we note that the longitudinal model used in these analyses compared individual trajectories over time (as opposed to differences between those who received and did not receive ART). Furthermore, as ART is freely available in our setting, socioeconomic barriers to accessing therapy would not have influenced our results . Second, although ART initiation was determined through a confidential linkage to a centralized dispensation database, all other behavioural data were self-reported. Although it is possible that stigmatized behaviours such as unprotected intercourse may have been underreported, we have no reason to believe that socially desirable reporting would have been more or less common during the visit following ART initiation. Thus, underreporting of these behaviours would have biased our results towards the null. Third, we were unable to assess changes in behaviour that occurred within 6 months of ART initiation due to the semi-annual design of the study. We were also unable to examine the potential role that individuals' perceptions regarding the ability of ART to reduce HIV transmissibility may have on return to engagement in risk behaviour following initiation. Finally, as ACCESS is not a random sample, we are unable to generalize our findings to the entire IDU population in Vancouver or to other settings; however, prior reports have indicated that this cohort's sociodemographics are similar to that of other samples of IDUs in Canada .
Targeting IDU populations in efforts to expand ART is increasingly recognized as a promising strategy to control HIV transmission and should be a key public health priority given the known impacts of ART in improving morbidity and mortality [30,31]. Concerns regarding the expansion of access to ART for IDU populations due to fears of increased HIV risk behaviour in the period following ART initiation are unsupported by the results presented in this study. Given these findings and the known benefits of ART among individuals with a history of injection drug use , we recommend the immediate implementation and evaluation of novel programmes to reduce barriers to HIV care and increase uptake of ART among IDUs.
The authors sincerely thank the study participants for their contribution to the research, as well as current and past researchers and staff. We would specifically like to thank Deborah Graham, Tricia Collingham, Caitlin Johnston, Steve Kain, and Calvin Lai for their research and administrative assistance.
Author contributions: E.W. had full access to all of the data in the study and takes full responsibility for the integrity of the data and the accuracy of the data analysis. B.D.L.M., M.-J.M., and E.W. designed the study and wrote the protocol. R.Z. conducted the statistical analysis and all authors interpreted the results. B.D.L.M. wrote the manuscript. M.-J.M., E.W., T.K., and J.S.G.M. critically revised the manuscript and contributed important intellectual content. All authors have read and approved the final version of the manuscript.
J.S.G.M. has received educational grants from and is serving as an ad hoc advisor to or speaking at various events sponsored by Abbott Laboratories, Agouron Pharmaceuticals Inc., Boehringer Ingelheim Pharmaceuticals Inc., Borean Pharma AS, Bristol-Myers Squibb, DuPont Pharma, Gilead Sciences, GlaxoSmithKline, Hoffmann-La Roche, Immune Response Corporation, Incyte, Janssen-Ortho Inc., Kucera Pharmaceutical Company, Merck Frosst Laboratories, Pfizer Canada Inc., Sanofi Pasteur, Shire Biochem Inc., Tibotec Pharmaceuticals Ltd, and Trimeris Inc. No other authors reported disclosures.
The study was supported by the US National Institutes of Health (R01DA021525) and the Canadian Institutes of Health Research (MOP-79297, RAA-79918). T.K. is supported by the Michael Smith Foundation for Health Research (MSFHR) and the Canadian Institutes of Health Research (CIHR). M.-J.M. is supported by a Doctoral Research Award from CIHR. B.D.L.M. is supported by a Doctoral Research Award from CIHR and a MSFHR Senior Trainee Award.
1. Hammer SM, Squires KE, Hughes MD, Grimes JM, Demeter LM, Currier JS, et al
. A controlled trial of two nucleoside analogues plus indinavir in persons with human immunodeficiency virus infection and CD4 cell counts of 200 per cubic millimeter or less. AIDS Clinical Trials Group 320 Study Team. N Engl J Med 1997; 337:725–733.
2. Hogg RS, Heath KV, Yip B, Craib KJ, O'Shaughnessy MV, Schechter MT, et al
. Improved survival among HIV-infected individuals following initiation of antiretroviral therapy. JAMA 1998; 279:450–454.
3. Gulick RM, Mellors JW, Havlir D, Eron JJ, Gonzalez C, McMahon D, et al
. Treatment with indinavir, zidovudine, and lamivudine in adults with human immunodeficiency virus infection and prior antiretroviral therapy. N Engl J Med 1997; 337:734–739.
4. Attia S, Egger M, Muller M, Zwahlen M, Low N. Sexual transmission of HIV according to viral load and antiretroviral therapy: systematic review and meta-analysis. AIDS 2009; 23:1397–1404.
5. Montaner JS, Hogg R, Wood E, Kerr T, Tyndall M, Levy AR, et al
. The case for expanding access to highly active antiretroviral therapy to curb the growth of the HIV epidemic. Lancet 2006; 368:531–536.
6. Kelly JA, Otto-Salaj LL, Sikkema KJ, Pinkerton SD, Bloom FR. Implications of HIV treatment advances for behavioral research on AIDS: protease inhibitors and new challenges in HIV secondary prevention. Health Psychol 1998; 17:310–319.
7. Crepaz N, Hart TA, Marks G. Highly active antiretroviral therapy and sexual risk behavior: a meta-analytic review. JAMA 2004; 292:224–236.
8. Ostrow DE, Fox KJ, Chmiel JS, Silvestre A, Visscher BR, Vanable PA, et al
. Attitudes towards highly active antiretroviral therapy are associated with sexual risk taking among HIV-infected and uninfected homosexual men. AIDS 2002; 16:775–780.
9. Tun W, Celentano DD, Vlahov D, Strathdee SA. Attitudes toward HIV treatments influence unsafe sexual and injection practices among injecting drug users. AIDS 2003; 17:1953–1962.
10. Dukers NH, Goudsmit J, de Wit JB, Prins M, Weverling GJ, Coutinho RA. Sexual risk behaviour relates to the virological and immunological improvements during highly active antiretroviral therapy in HIV-1 infection. AIDS 2001; 15:369–378.
11. Miller M, Meyer L, Boufassa F, Persoz A, Sarr A, Robain M, et al
. Sexual behavior changes and protease inhibitor therapy. SEROCO Study Group. AIDS 2000; 14:F33–F39.
12. Scheer S, Kellogg T, Klausner JD, Schwarcz S, Colfax G, Bernstein K, et al
. HIV is hyperendemic among men who have sex with men in San Francisco: 10-year trends in HIV incidence, HIV prevalence, sexually transmitted infections and sexual risk behaviour. Sex Transm Infect 2008; 84:493–498.
13. World Health Organization (WHO). Antiretroviral therapy and injecting drug users.
In: Evidence for action series
. Geneva: WHO; 2005.
14. Mathers BM, Degenhardt L, Ali H, Wiessing L, Hickman M, Mattick RP, et al
. HIV prevention, treatment, and care services for people who inject drugs: a systematic review of global, regional, and national coverage. Lancet 2010; 375:1014–1028.
15. Battegay M, Bucher HC, Vernazza P. Sexual risk behavior in HIV-infected injection drug users. Clin Infect Dis 2004; 38:1175–1177.
16. Tun W, Gange SJ, Vlahov D, Strathdee SA, Celentano DD. Increase in sexual risk behavior associated with immunologic response to highly active antiretroviral therapy among HIV-infected injection drug users. Clin Infect Dis 2004; 38:1167–1174.
17. Vlahov D, Safaien M, Lai S, Strathdee SA, Johnson L, Sterling T, et al
. Sexual and drug risk-related behaviours after initiating highly active antiretroviral therapy among injection drug users. AIDS 2001; 15:2311–2316.
18. Bouhnik AD, Moatti JP, Vlahov D, Gallais H, Dellamonica P, Obadia Y. Highly active antiretroviral treatment does not increase sexual risk behaviour among French HIV infected injecting drug users. J Epidemiol Community Health 2002; 56:349–353.
19. Wood E, Kerr T, Hogg RS, Zhang R, Tyndall MW, Montaner JS. Validity of self-reported antiretroviral therapy use among injection drug users. J Acquir Immune Defic Syndr 2006; 41:530–531.
20. Strathdee SA, Palepu A, Cornelisse PG, Yip B, O'Shaughnessy MV, Montaner JS, et al
. Barriers to use of free antiretroviral therapy in injection drug users. JAMA 1998; 280:547–549.
21. Wood E, Hogg RS, Bonner S, Kerr T, Li K, Palepu A, et al
. Staging for antiretroviral therapy among HIV-infected drug users. JAMA 2004; 292:1175–1177.
22. Fitzmaurice GM, Laird NM, Ware JH. Applied longitudinal analysis. Hoboken, New Jersey: Wiley; 2004.
23. Vittinghoff E, Glidden DV, Shiboski SC, McCulloch CE. Regression models in biostatistics: linear, logistic, survival, and repeated measures models. New York: Springer; 2005.
24. Rothman KJ, Greeland S, Lash TL. Modern epidemiology. 3rd ed Philadelphia, Pennsylvania: Lippincott Williams & Wilkins; 2008.
25. Rhodes T, Singer M, Bourgois P, Friedman SR, Strathdee SA. The social structural production of HIV risk among injecting drug users. Soc Sci Med 2005; 61:1026–1044.
26. Wolitski RJ, Kidder DP, Fenton KA. HIV, homelessness, and public health: critical issues and a call for increased action. AIDS Behav 2007; 11:S167–S171.
27. Galea S, Vlahov D. Social determinants and the health of drug users: socioeconomic status, homelessness, and incarceration. Public Health Rep 2002; 117:S135–S145.
28. Carrieri MP, Moatti JP, Vlahov D, Obadia Y, Reynaud-Maurupt C, Chesney M. Access to antiretroviral treatment among French HIV infected injection drug users: the influence of continued drug use. MANIF 2000 Study Group. J Epidemiol Community Health 1999; 53:4–8.
29. Public Health Agency of Canada (PHAC). I-Track: enhanced surveillance of risk behaviours among people who inject drugs. Phase I Report, August 2006
. Ottawa: Surveillance and Risk Assessment Division, Centre for Infectious Disease Prevention and Control, PHAC; 2006.
30. Wood E, Montaner JS, Bangsberg DR, Tyndall MW, Strathdee SA, O'Shaughnessy MV, et al
. Expanding access to HIV antiretroviral therapy among marginalized populations in the developed world. AIDS 2003; 17:2419–2427.
31. Donoghoe MC, Bollerup AR, Lazarus JV, Nielsen S, Matic S. Access to highly active antiretroviral therapy (HAART) for injecting drug users in the WHO European Region 2002-2004. Int J Drug Policy 2007; 18:271–280.
32. Wood E, Hogg RS, Lima VD, Kerr T, Yip B, Marshall BD, et al
. Highly active antiretroviral therapy and survival in HIV-infected injection drug users. JAMA 2008; 300:550–554.