Injection drug use continues to be a major risk factor for acquisition of HIV and hepatitis viruses (1–8). Studies among injection drug users (IDUs) have shown that young (e.g., ≤21 or ≤35 years of age) or recently initiated (e.g., injecting ≤2 or ≤5 years) IDUs are at an increased risk for HIV infection (1,2,9–14). Among the few studies conducted among young or recently initiated IDUs (or “new IDUs”), both high-risk sexual practices (e.g., trading sex, multiple sex partners, male-with-male sex) and injection practices (e.g., injecting cocaine/speedball, sharing injection equipment with high-risk network) have been found to be associated with HIV infection (15,16). From these studies, researchers have emphasized the need to study new IDUs to better understand high-risk practices and circumstances (3,15,17–19).
In a study that examined the prevalence of four blood-borne infections, Garfein and colleagues (5) noted that even within the first 2 years of starting injection drug use, rates of hepatitis B virus (HBV) and hepatitis C virus (HCV) infection exceeded 50%. Follow-up studies of new IDUs revealed a high incidence of infection within the first 3 years of starting injection (1,2,4,6,10,16). Thus, further characterization of the high-risk period of initiation of injection drug use may be useful in designing intervention programs aimed at reducing risk within this hidden population, because current harm reduction strategies (i.e., syringe exchange), particularly in Baltimore City, have fallen short (20). One long-standing risk factor for HIV infection among adult IDUs is attending a shooting gallery (SG)—a hidden location within a drug-using neighborhood, where multiperson use of injection equipment commonly occurs (21–29). It has been hypothesized that new IDUs may only inject recreationally and are less likely to be part of the street drug scene (30). If this is the case, young persons who begin injecting might not frequent SGs or may be less aware of the HIV risks commonly experienced in these settings if they do frequent such settings. Although recent evidence from New York City suggests that young transient IDUs who attend SGs carry their own “new” injection equipment (15), SGs should nevertheless be considered as high-risk settings for acquiring HIV and other blood-borne infections, because repackaging and selling of used injection equipment continues to take place in these settings (31).
Herein, we report on circumstances associated with early (e.g., within the first year of initiating injection) SG attendance as well as HIV prevalence and incidence rates among 226 adolescent and young adult IDUs who were recruited into a prospective study. At enrollment, we conducted year-by-year risk histories of sex and drug use practices spanning the first year of initiating injection drug use to the year of study entry. An earlier analysis identified early SG attendance as risk factor for HIV prevalence among this population of new IDUs (32). Thus, the purpose of this study was to examine those new IDUs who injected in a SG early on in their injection career. We specifically examined early sex practices, injection practices, and high-risk social circumstances that occurred within the first year after initiation, given the early high risk for HIV and HCV transmission after the start of an injecting career. We hypothesize that early SG attendees are characterized by other high-risk social circumstances, particularly those related to initiating injection drug use, after taking other high-risk injection practices into account. These analyses were conducted to help determine if new IDUs who begin high-risk practices early can be characterized by other high-risk circumstances that can be used to design harm reduction approaches among this hidden vulnerable population.
MATERIALS AND METHODS
Between July 1997 and May 1999, adolescent and young adult IDUs were recruited into a prospective study of HIV infection in Baltimore (Risk Evaluation and Assessment of Community Health [REACH II] Study). A recreational vehicle (mobile clinic) was used to enhance accessibility to this hidden population. A stationary clinic site also operated daily in the East Baltimore area of the city in conjunction with the REACH II Study mobile clinic. Experienced street outreach workers who engaged young drug users on the street served as the primary method of recruitment. The outreach team also posted flyers in various settings where young IDUs might attend such as youth shelters, treatment centers, emergency rooms, public health clinics, the Baltimore Needle Exchange Program, and neighborhoods of vacant buildings. Participants were also recruited through word of mouth.
To be considered eligible for the study, persons were required to be 15 to 30 years of age (verified by photo identification), have injected ≤5 years prior to study entry, and have reported at least one injection in the 6 months prior to study entry. Injection status was verified by presence of stigmata or “track marks.” Participants who injected for less than 1 year and less than once per week were not required to have track marks to gain entry into the study. At each scheduled study visit, a modest remuneration ($20) was provided. During this study, non-IDUs (crack smokers and intranasal heroin/cocaine users) were enrolled simultaneously to reduce the incentive among drug users to fabricate an injection history to gain study entry. The Institutional Review Board of the Johns Hopkins Bloomberg School of Public Health (Committee on Human Research) approved this study.
Data collection included screening for eligibility (by the outreach worker during the recruitment phase and by the interviewer during the enrollment phase), followed by informed consent and a structured interview conducted privately by a trained interviewer. Venipuncture and pretest HIV counseling were conducted after the interview. Participants were informed at the beginning of the interview that they would see a counselor immediately after the interview, at which time they would receive referrals for medical and social services. A 2-week HIV test result visit was scheduled, during which time the participant received additional counseling and referrals. This study also included a 6-month follow-up visit and a 12-month follow-up visit, which consisted of an interviewer-administered questionnaire and venipuncture to permit an estimate of HIV incidence.
Demographic characteristics examined in this study included age, sex, race/ethnicity, education, and juvenile arrest. Early sex practice, which included trading sex, was examined using a reconstruction of behavioral histories. This history was conducted with a retrospective year-by-year assessment spanning the year of injection initiation to study entry using an interviewer strategy described previously (33,34). This history began with behaviors occurring in the most remote period and moved forward in time to the most recent period. Other early sex practices explored included ≤13 years of age at first sexual experience and same-sex practice (e.g., male with male, female with female); however, because too few male IDUs reported this behavior, we were unable to sufficiently explore this risk practice.
Early injection practices that occurred during the first year of initiating injection drug use included frequency of injection, sharing injection equipment, and injecting within a high-risk network. The SG attendance outcome variable was defined as those reporting visiting a SG and using needles, cotton, or cookers borrowed, rented, or bought from there, which were behaviors captured through the year-by-year history. Other injection exposure variables included age at initiation of injection drug use, type of drug used at initiation of injection (e.g., cocaine/speedball vs. heroin only), and duration of injection use.
Social circumstances of initiation of injection drug use associated with early SG attendance included the relationship the new IDU had with the person who introduced injection drug use (“initiator”) to him or her, age of the initiator, and whether the new IDU had ever introduced someone to injection drug use since he or she had begun injecting.
At each study visit, a phlebotomist drew 10 mL of whole blood, which was transported daily to the Department of Epidemiology's Infectious Disease Program laboratory for HIV antibody detection. HIV antibody was detected by ELISA (Ortho Diagnostics, Raritan, NJ, U.S.A.) and confirmed with the Western blot test (WB) (Ortho Diagnostics) by trained laboratory personnel using standard criteria.
Variables explored as potential correlates of early SG attendance included demographic and serologic characteristics. Race/ethnicity was categorized as African American compared with white or other race, with less than 5% identifying as other race/ethnicity. Race/ethnicity was combined in this fashion because of the increased risk of HIV among African-American IDUs and also so that a similar comparison across previous studies could be made (28). Education was defined as having less than a high school diploma compared with having at least a high school diploma or GED.
In terms of early injection practices, injecting more often than once per day, sharing injection equipment, and injecting within a high-risk network during the first year of initiating injection drug use were examined. An injecting network was considered to be high risk if an IDU injected with ≥2 other IDUs and at least 1 of those persons was not personally known by the IDU. Other injection practices included duration of injection drug use (≤2 vs. 2–5 years), drug type injected at initiation (cocaine or speedball vs. heroin only), and initiating injection drug use during adolescence versus young adulthood (≤21 vs. >21 years of age). Early high-risk sex practices included trading sex during the first year of initiating injection drug use (ever vs. never), age of first sexual experience (≤13 vs. >13 years of age), and sexual identity (gay/bisexual male vs. heterosexual male, and lesbian/bisexual female vs. heterosexual female). Because of sample size constraints, however, sexual identity was not included in the final model-building stage of the analysis. Social circumstances of initiation of injection drug use included “who” initiated the IDU into injection drug use (e.g., self, sex partner, friend, relative), age of initiator relative to the IDU (e.g., ≥5 vs. <5 years older than the IDU), and subsequently introducing someone else into injection drug use (ever vs. never).
To compare high-risk injection, sex practices, and social circumstances of the first injection that was associated with early SG attendance, simple descriptive statistics were performed for all variables of interest. In the case where continuous data were skewed, medians and ranges were used as opposed to variable means. Bivariate analyses examining exposure variables by early SG attendance were conducted using Fisher exact tests and Mantel-Haenszel ORs. Significant bivariate associations using p values and 95% CIs were used to assist with determining which exposure variables to explore further during model building. A final multiple logistic regression model was developed to determine the independent effect of each variable of interest as it related to early SG attendance.
Date of HIV seroconversion was estimated to occur at the midpoint between the last seronegative visit and the first seropositive visit. Person-time analysis was used to estimate HIV incidence among REACH II Study cohort members who returned for follow-up. CIs for the incidence estimate were based on the Poisson distribution.
Of the 276 IDUs screened, 226 (82%) were eligible for the study according to the study criteria and 75 (33%) reported early SG attendance. The median age was 25 years, 64% were African American, and 61% were female (Table 1). Over half of the participants had dropped out of high school and had not obtained a GED (62%), one fifth had experienced a juvenile arrest (20%), and 24 participants tested seropositive for HIV antibody (11%). Early SG attendees were nearly three times more likely to be HIV-seroprevalent (OR = 2.7) than nonattendees.
In terms of early injection practices, the median age of initiating injection drug use was 23 years (range: 10–30 years), with 61% reporting their first injection ≤2 years prior to study entry (Table 2). The majority injected heroin at initiation (81%), and about half reported frequent injection (45%). During the year after initiation of injection drug use, sharing injection equipment (OR = 5.0) and injecting within a high-risk network (OR = 3.3) were the injection practices significantly associated with early SG attendance.
With respect to sex practices, 12% of the total study population self-identified as homosexual or bisexual (3% of male participants and 18% of female participants). Early SG attendance was significantly higher among female participants who identified themselves as being lesbian or bisexual (OR = 3.3) compared with those who did not. Only three male participants identified themselves as being gay or bisexual, none of whom reported early SG attendance. Compared with those who did not report early SG attendance, early sex trading was significantly higher among the early SG attendees (OR = 2.2).
Table 3 depicts social circumstances of initiation of injection drug use associated with early SG attendance. In terms of the relationship with the person the IDU was initiated by, early SG attendees were significantly more likely to be initiated by a friend (OR = 1.9) compared with nonattendees. In terms of the age of the initiator, early SG attendees were significantly more likely to be initiated by someone ≥5 years their age (OR = 2.2) compared with nonattendees. There was also a tendency for early SG attendees to introduce someone to injection drug use subsequently (OR = 1.9) compared with nonattendees.
In Table 4, we examined the relation between early SG attendance and high-risk practices and social circumstances of initiation of injection drug use. Using multiple logistic regression, early SG attendees were more likely to be HIV-seropositive (adjusted OR [AOR] = 3.1) and to share syringes during the first year after initiation of injection drug use (AOR = 5.0) compared with early SG nonattendees. Within the same final model, early SG attendees were also more likely to be initiated by someone ≥5 years their age (AOR = 2.0) compared with early SG nonattendees.
Of the 226 new IDUs studied at baseline, 50% returned for at least one follow-up visit over a 12-month period, with 111.8 person-years of follow-up. There were six HIV seroconversions observed, yielding an annual incidence of 6.6 per 100 person-years (95% CI: 2.20–13.30). The small number of seroconversions observed precluded further analyses to examine potential risk factors. There were no demographic differences with respect to those who returned for follow-up and those who did not.
A major finding in this study was that new IDUs in Baltimore are in need of increased harm reduction strategies given their high incidence of HIV infection and lack of access to current harm reduction services (20). To date, this is the highest HIV incidence rate observed among young or recent IDUs (35). Although the small number of seroconversions prevented further analysis of risk factors, HIV prevalence was significantly associated with early SG attendance in this population of young recent initiates. SG attendance has been well established as a risk factor for HIV infection among adult IDUs (21–29); however, to our knowledge, this is the first study to identify an independent association between HIV seroprevalence and early SG attendance among new IDUs. In fact, HIV seroprevalence demonstrated a stronger association with early SG attendance (OR = 3.1) than did HIV seroprevalence with any SG attendance (OR = 2.2), which supports previous findings that the first year after initiation of injection is a high-risk period (3,5,6). Evidence of this high-risk period is further supported by the fact that 80% of those who reported any SG attendance reported use within the first year of starting to inject. Earlier studies that have identified SG use as a risk factor for HIV measured attendance over a broad (e.g., lifetime) (17) or recent period (e.g., past 6 months) (36). In this study, SG attendance not only measured a specific high-risk period (e.g., year after first injection) but also measured “attending a SG and using drug paraphernalia that had been bought, rented, or borrowed from there.” This refined measure could account for the high proportion of HIV prevalence observed among this new IDU population of early SG attendees.
A second major finding in this study was that after adjusting for high-risk injection practices (e.g., early injection equipment sharing), early SG attendees were more likely to be introduced to injection drug use by an older IDU peer compared with early SG nonattendees. Previous research has shown that those who inject with older IDUs are at an increased risk for HCV (6). Additionally, our findings also revealed early SG attendees to inject within high-risk networks, which is not entirely surprising, because SGs are often places where many IDUs who may commonly be strangers to one another gather to inject drugs. Thus, not only are new IDUs sharing and injecting with multiple persons, some of whom they do not personally know, but their level of risk is potentially increased when introduced to injection drug use by an older IDU, due to the higher HIV prevalence pool among older, more established IDUs. A previous study conducted among SG attendees in Puerto Rico found that the long-term SG managers were significantly more likely to be HIV-seroprevalent than short-term managers (37), which provides further evidence of the potential risk of being introduced to injection by an older IDU peer. Although it has been suggested that young transient IDUs in New York City tend to carry their own “new” needles to SGs (15), such behavior has not been observed among young IDUs in Baltimore. Yet, a recent study conducted among adult IDUs in Baltimore found that SG attendees were more likely to report using the needle exchange program (NEP) than nonattendees (38). Similar findings were also observed among adult female IDUs in Canada (39). These findings could suggest that the proportion of sterile syringes in SGs might have increased since NEPs have opened. However, it is unclear through this study and previous studies if the sterile syringes from NEPs are new when they are carried into SGs or if they are being repackaged and sold as new after they have been previously used. Quantitative and ethnographic research on SG settings is necessary to get a sense of the type of high-risk behavior that occurs within these settings (e.g., nonuse or lack of water or bleach, peer pressure to share, lack of HIV disclosure) as well as the accessibility to sterile syringes within these settings across the United States and abroad. This potentially high-risk circumstance emphasizes the importance of social network–based interventions within SG drug networks.
There are some limitations to this study. Because sampling was based on self-referral, certain subgroups might have had differential levels of comfort in approaching and being recruited into the study (i.e., drug-injecting subnetworks of young women as evidenced by the relatively large proportion enrolled in the study). To reduce this potential selection bias, we used a mobile clinic to travel to several different high-risk neighborhoods in Baltimore City to broaden recruitment.
In cross-sectional studies, misclassification of exposure can occur as a result of biased recall or difficulty in recalling behavioral events. For example, some HIV-seropositive participants may have had prior knowledge of their serostatus, enabling them to remember events surrounding injection initiation with greater accuracy compared with those with negative serostatus, who lack a similar memorable event. To minimize recall bias, we referred to other memorable biographic landmarks (i.e., pregnancy, childbirth, time spent incarcerated, last school grade attended) that could assist HIV-seronegative participants in recalling the chronologic sequence of high-risk behaviors. Furthermore, by collapsing the years during the analysis, the potential for this type of inaccuracy was reduced. In spite of these measures taken, nondifferential misclassification could have occurred if a participant could not accurately remember behavioral events during the year-by-year history (e.g., “telescoping” and “distancing”) (34), which would tend to bias associations toward the null.
This study had limited power to detect differences between specific risk behaviors. For example, because of small sample sizes or underreporting of gay or bisexual male participants and lesbian or bisexual female participants, we were unable to explore the association between early SG attendance and gender-specific sexual identity. Given the increased risk associated with same-sex sexual identity among males and females, particularly with respect to injection practices (5,29,36,39–42), further investigation is needed among same-sex sexual-identifying, new IDUs.
It should also be noted that the comparison group in this analysis actually contained those who never attended SGs and later SG attendees. Given the known association between HIV and SG use (regardless if it is early or late in an injecting career), our estimates reported here are likely to be underestimated.
Despite these limitations, we have identified high-risk social circumstances of initiation of injection drug use among early SG attendees, which, in turn, highlight the importance of peer-driven or network-oriented interventions. For example, a new IDU being initiated into injection drug use by an older IDU (with approximately 85% reporting injection initiation at home or at a friend's home) and subsequently injecting within a high-risk social setting (e.g., a high-risk network and in a shooting gallery) suggests that these younger early SG attendees may have a diverse and perhaps unstable network. Thus, these individuals may be a bridge group, passing HIV from older IDUs to new IDUs. Intervention strategies should be aimed at all SG attendees who can serve as peer educators within this high-risk drug network. In doing this, targeted efforts will reach high-risk, young, new IDUs, who are typically difficult to reach and identify. Because network-oriented intervention efforts that have used IDU peer educators have proven to be successful among adult populations (31,43), similar strategies should be expanded to young recent IDUs as well. It is also apparent from these findings that SG settings should be considered practical venues when designing intervention strategies. Earlier studies involving the use of SGs has proven to be feasible and successful (31,44–46). In addition, social network analyses could be used to identify key individuals within SGs to refine these targeted efforts.
Future research should include quantitative and ethnographic studies of SGs so that intervention strategies can be carefully tailored to specific neighborhoods and IDU communities, because we can only generalize our findings to young new IDUs in Baltimore. Determining why and under what circumstances a new IDU would inject in a SG so soon after initiating injection drug use may be useful for designing programs aimed at this high-risk subgroup. Based on previous reports, socioeconomic instability could play a role in early SG attendance among younger IDUs just as it has with adult IDUs (37,47). Other possible explanations could be that young new IDU SG attendees may be part of an older IDU drug network or lack a network altogether. Exploring these possible explanations among young and older IDUs would further assist with the planning and implementation of effective social network prevention and intervention programs.
The authors thank the REACH II Study team and all the staff at the Ferne Johnson Center. Ferne Johnson was a Community Advisory Board member and was one of the first activists in East Baltimore to dedicate her life to those struggling with HIV disease, particularly injection drug users, early on in the epidemic. The authors are especially grateful to the young participants of the REACH II Study for sharing their experiences with us and to the Community Advisory Board members without whom this research would not have been possible.
1. Nelson KE, Vlahov D, Solomon L, et al. Temporal trends of incident human immunodeficiency virus infection in a cohort of injection drug users
in Baltimore, Maryland. Arch Intern Med 1995; 155:1305–11.
2. Nicolosi A, Leite MLC, Molinari S, et al. Incidence and prevalence trends of HIV
infection in intravenous drug users attending treatment centers in Milan and Northern Italy, 1986–1990. J Acquir Immune Defic Syndr Hum Retrovirol 1992; 5:365–73.
3. Vlahov D, Anthony JC, Celentano DD, et al. Trend of HIV
-1 risk reduction among initiates into intravenous drug use 1982–1987. Am J Drug Alcohol Abuse 1991; 17:39–48.
4. van Ameijden EJC, van den Hoek JAR, Mientjes GHC, et al. A longitudinal study on the incidence and patterns of HIV
, HBV, and HCV infection among drug users in Amsterdam. Eur J Epidemiol 1993; 9:255–62.
5. Garfein RS, Vlahov D, Galai N, et al. Viral infections in short-term injection drug users
: the prevalence of the hepatitis C, hepatitis B, human immunodeficiency virus, and human T-lymphotropic viruses. Am J Public Health 1996; 86:655–61.
6. Garfein RS, Doherty MC, Monterroso ER, et al. Prevalence and incidence of hepatitis C virus infection among young
adult injection drug users
. J Acquir Immune Defic Syndr Hum Retrovirol 1998; 18(Suppl):S11–9.
7. Thomas DL, Vlahov D, Solomon L, et al. Correlates of hepatitis C virus infection among injection drug users
. Medicine 1995; 74:212–20.
8. Villano SA, Vlahov D, Nelson KE, et al. Incidence and risk factors for hepatitis C among injection drug users
in Baltimore, Maryland. J Clin Microbiol 1997; 35:3274–7.
9. van Ameijden EJC, van den Hoek JAR, van Haastrecht HJA, et al. The harm reduction approach and risk factors for human immunodeficiency virus (HIV
) seroconversion in injecting drug users, Amsterdam. Am J Epidemiol 1992; 136:236–43.
10. Fennema JSA, van Ameijden EJC, van den Hoek A, et al. Young
and recent-onset injecting drug users are at higher risk for HIV
. Addiction 1997; 92:1457–65.
11. Booth R, Koester S, Brewster JT, et al. Intravenous drug users and AIDS: risk behaviors. Am J Drug Alcohol Abuse 1991; 17:337–53.
12. Marrero Rodriguez CA, Robles RR, Colon HM, et al. HIV
risk behaviors and HIV
seropositivity among young injection drug users
. PR Health Sci J 1993; 12:7–12.
13. Telles PR, Bastos FI, Guydish J, et al. Risk behavior and HIV
seroprevalence among injecting drug users in Rio de Janeiro, Brazil. AIDS 1997;(Suppl 1):S35–42.
14. Carneiro M, Fuller C, Doherty MC, et al. HIV
prevalence and risk behaviors among new initiates into injection drug use over the age of 40 years old. Drug Alcohol Depend 1999; 54:83–6.
15. Des Jarlais DC, Friedman SR, Perlis T, et al. Risk behavior and HIV
infection among new drug injectors in the era of AIDS in New York City. J Acquir Immune Defic Syndr 1999; 20:67–72.
16. Doherty MC, Garfein RS, Monterroso EM, et al. Correlates of HIV
infection among young
adult short-term injection drug users
. AIDS 2000; 14:717–26.
17. Battjes RJ, Leukefeld CG, Pickens RW. Age at first injection and HIV
risk among intravenous drug users. Am J Drug Abuse 1992; 18:263–73.
18. Roy E, Haley N, Boivin JY, et al. Predictors of drug injection initiation in a street youth cohort [abstract 33389]. Presented at the 12th World AIDS Conference, Geneva, June 1998.
19. Diaz T, Conover S, Edwards V, et al. Drug using behaviors of young
and recent initiate injection drug users
in New York City: a unique opportunity for prevention of HIV
[abstract 23228]. Presented at the 12th World AIDS Conference, Geneva, June 1998.
20. Valente TW, Foreman RK, Junge B, et al. Needle-exchange participation, effectiveness, and policy: syringe relay, gender, and the paradox of public health. J Urban Health 2001; 78:340–9.
21. Marmor M, Des Jarlais DC, Cohen H, et al. Risk factors for infection with human immunodeficiency virus among intravenous drug abusers in New York City. AIDS 1987; 1:39–44.
22. Schoenbaum EE, Hartel D, Selwyn PA. Risk factors for human immunodeficiency virus infection in intravenous drug users. N Engl J Med 1989; 321:874–9.
23. Chaisson RE, Bacchetti P, Osmond D, et al. Cocaine use and HIV
infection in intravenous drug users in San Francisco. JAMA 1989; 26:561–5.
24. Vlahov D, Munoz A, Anthony JC, et al. Association of drug injection patterns with antibody to human immunodeficiency virus type 1 among intravenous drug users in Baltimore, Maryland. Am J Epidemiol 1990; 132:847–56.
25. Koblin BA, McCusker J, Lewis BJ, et al. Racial/ethnic differences in HIV
-I seroprevalence and risky behaviors among intravenous drug users in a multisite study. Am J Epidemiol 1990; 132:837–46.
26. Nelson KE, Vlahov D, Cohn S, et al. Sexually transmitted diseases in a population of intravenous drug users: association with seropositivity to the human immunodeficiency virus (HIV
). J Infect Dis 1991; 164:457–63.
27. Moss AR, Vranizan K, Gorter R, et al. HIV
seroconversion in intravenous drug users in San Francisco, 1985–1990. AIDS 1994; 8:223–31.
28. Battjes RJ, Pickens RW, Haverkos HW, et al. HIV
risk factors among injecting drug users in five US cities. AIDS 1994; 8:681–7.
29. Strathdee SA, Galai N, Safaiean M, et al. Sex differences in risk factors for HIV
seroconversion among injection drug users
: a ten year perspective. Arch Int Med 2001: 161:1281–8.
30. Neaigus A, Friedman SR, Jose B, et al. High-risk personal networks and syringe sharing as risk factors for HIV
infection among new drug injectors. J Acquir Immune Defic Syndr Hum Retrovirol 1996; 11:499–509.
31. Page BJ. Shooting scenarios and risk of HIV
-1 infection. Am Behav Sci 1990; 33:478–90.
32. Fuller C, Vlahov D, Ompad DC, et al. Correlates of HIV
infection among newly initiated adolescent and young
adult injection drug users
(REACH II) [abstract]. Presented at the 32nd Annual Meeting Society for Epidemiologic Research, Baltimore, 1999.
33. Anthony JC, Vlahov D, Celentano DD, et al. Self-report interview data for study of HIV
-1 infection among intravenous drug users: description of methods and preliminary evidence on validity. J Drug Issues 1991; 21:739–57.
34. Samuels JF, Vlahov D, Anthony JC, et al. Measurement of HIV
risk behaviors among intravenous drug users. Br J Addict 1992; 87:417–28.
35. Monterroso ER, Hamburger ME, Vlahov D, et al. Prevention of HIV
infection in street-recruited injection drug users
. J Acquir Immune Defic Syndr 2000; 25:63–70.
36. Celentano D, Vlahov D, Cohn S, et al. Risk factors for shooting gallery
use and cessation among intravenous drug users. Am J Public Health 1991; 81:1291–5.
37. Robles RR, Marrero CA, Reyes JC, et al. J Acquir Immune Defic Syndr Hum Retrovirol 1998; 17:477–83.
38. Latkin CA, Forman Valerie L. Patterns of needle acquisition and sociobehavioral correlates of needle exchange program attendance in Baltimore, Maryland, U.S.A. J Acquir Immune Defic Syndr 2001; 27:398–404.
39. Archibald CP, Ofner M, Strathdee SA, et al. Factors associated with frequent needle exchange program attendance in injection drug users
in Vancouver, Canada. J Acquir Immune Defic Syndr Hum Retrovirol 1998; 17:160–6.
40. Reardon J, Wilson MJ, Lemp GF, et al. HIV
-1 infection among female injection drug users
(IDU) in the San Francisco Bay Area, California, 1989–1991 [abstract Th.C.1553]. Presented at the VIII International Conference on AIDS, Amsterdam, July 1992.
41. Friedman SR, Des Jarlais DC, Deren S, et al. HIV
seroconversion among female drug injectors who have sex with women [abstract]. Presented at the 120th Annual Meeting of the American Public Health Association, New York, 1992.
42. Magura S, Kang S, Shapiro J, et al. HIV
risk among women injecting drug users who are in jail. Addiction 1993; 88:1351–60.
43. Latkin C. Outreach in natural settings: The use of peer leaders for HIV
prevention among injection drug users
' networks. Public Health Rep 1998; 113(Suppl):151–9.
44. Chitwood DD, McCoy CB, Inciardi JA, et al. HIV
seropositivity of needles from shooting galleries in South Florida. Am J Public Health 1990; 80:150–2.
45. Shah SM, Shapshak P, Rivers JE, et al. Detection of HIV
-1 DNA in needle/syringes, paraphernalia, and washes from shooting galleries in Miami: a preliminary laboratory report. J Acquir Immune Defic Syndr Hum Retrovirol 1996; 11:301–6.
46. Shapshak P, Fujimura RK, Page JB, et al. HIV
-1 RNA load in needles/syringes from shooting galleries in Miami: a preliminary laboratory report. Drug Alcohol Depend 2000; 58:153–7.
47. Latkin CA, Mandell W, Vlahov D, et al. Personal network characteristics as antecedents to patterns of needle sharing and shooting gallery
attendance. Social Networks