Integrated respondent-driven sampling and peer support for persons who inject drugs in Haiphong, Vietnam: a case study with implications for interventions.

Combined prevention for HIV among persons who inject drugs (PWID) has led to greatly reduced HIV transmission among PWID in many high-income settings, but these successes have not yet been replicated in resource-limited settings. Haiphong, Vietnam experienced a large HIV epidemic among PWID, with 68% prevalence in 2006. Haiphong has implemented needle/syringe programs, methadone maintenance treatment (MMT), and anti-retroviral treatment (ART), but there is an urgent need to identify high-risk PWID and link them to services. We examined integration of respondent-driven sampling (RDS) and strong peer support groups as a mechanism for identifying high-risk PWID and linking them to services. The peer support staff performed the key tasks that required building and maintaining trust with the participants, including recruiting the RDS seeds, greeting and registering participants at the research site, taking electronic copies of participant fingerprints (to prevent multiple participation in the study), and conducting urinalyses. A 6-month cohort study with 250 participants followed the RDS cross-sectional study. The peer support staff maintained contact with these participants, tracking them if they missed appointments, and providing assistance in accessing methadone and ART. The RDS recruitment was quite rapid, with 603 participants recruited in three weeks. HIV prevalence was 25%, Hepatitis C (HCV) prevalence 67%, and participants reported an average of 2.7 heroin injections per day. Retention in the cohort study was high, with 86% of participants re-interviewed at 6-month follow-up. Assistance in accessing services led to half of the participants in need of methadone enrolled in methadone clinics, and half of HIV-positive participants in need of ART enrolled in HIV clinics by the 6-month follow-up. This study suggests that integrating large-scale RDS and strong peer support may provide a method for rapidly linking high-risk PWID to combined prevention and care, and greatly reducing HIV transmission among PWID in resource-limited settings.

Simultaneous recruitment of drug users and men who have sex with men in the United States and Russia using respondent-driven sampling: sampling methods and implications.

The Sexual Acquisition and Transmission of HIV Cooperative Agreement Program (SATHCAP) examined the role of drug use in the sexual transmission of the human immunodeficiency virus (HIV) from traditional high-risk groups, such as men who have sex with men (MSM) and drug users (DU), to lower risk groups in three US cities and in St. Petersburg, Russia. SATHCAP employed respondent-driven sampling (RDS) and a dual high-risk group sampling approach that relied on peer recruitment for a combined, overlapping sample of MSM and DU. The goal of the sampling approach was to recruit an RDS sample of MSM, DU, and individuals who were both MSM and DU (MSM/DU), as well as a sample of sex partners of MSM, DU, and MSM/DU and sex partners of sex partners. The approach efficiently yielded a sample of 8,355 participants, including sex partners, across all four sites. At the US sites-Los Angeles, Chicago, and Raleigh-Durham-the sample consisted of older (mean age = 41 years), primarily black MSM and DU (both injecting and non-injecting); in St. Petersburg, the sample consisted of primarily younger (mean age = 28 years) MSM and DU (injecting). The US sites recruited a large proportion of men who have sex with men and with women, an important group with high potential for establishing a generalized HIV epidemic involving women. The advantage of using the dual high-risk group approach and RDS was, for the most part, the large, efficiently recruited samples of MSM, DU, and MSM/DU. The disadvantages were a recruitment bias by race/ethnicity and income status (at the US sites) and under-enrollment of MSM samples because of short recruitment chains (at the Russian site).

HIV infection, sexual risk behavior, and substance use among Latino gay and bisexual men and transgender persons.

OBJECTIVES: We examined HIV prevalence and the socioeconomic correlates of HIV infection, sexual risk behaviors, and substance use among Latino gay and bisexual men and transgender persons in Chicago and San Francisco. METHODS: Data were collected from a sample of 643 individuals (Chicago: n=320; San Francisco: n=323) through respondent-driven sampling and computer-assisted self-administered interviews. RESULTS: HIV prevalence in San Francisco (0.325; 95% confidence interval [CI]=0.260, 0.393) was higher than in Chicago (0.112; 95% CI=0.079, 0.163). In San Francisco, HIV prevalence was higher among US-born residents than among those born outside the country; in Chicago, the opposite was true. Heavy use of alcohol was prevalent, especially in Chicago (0.368; 95% CI=0.309, 0.432; San Francisco: 0.154; 95% CI=0.116, 0.192). Drug use and more education were positively correlated and greater age was negatively correlated with unprotected anal intercourse. CONCLUSIONS: Heavy alcohol drinking and use of drugs remain a significant public health problem in this population. Drug use was more closely linked to HIV sexual risk behaviors than was heavy drinking.

Using dual capture/recapture studies to estimate the population size of persons who inject drugs (PWID) in the city of Hai Phong, Vietnam.

BACKGROUND: Good estimates of key population sizes are critical for appropriating resources to prevent HIV infection. We conducted two capture/recapture studies to estimate the number of PWID currently in Hai Phong, Vietnam. METHODS: A 2014 respondent-driven sampling (RDS) survey served as one capture, and distribution of cigarette lighters at drug use "hotspots" in 2016 served as another "capture." A 2016 survey using RDS, conducted 1 week after lighter distribution, served as "recapture" for both captures. Recaptured participants in the two surveys were identified with a computerized fingerprint reader. Recaptured participants from the lighter distribution were asked to show their lighters. RESULTS: 1385 participants were included in the "recapture" survey. They were 94% male and had a median age of 39. All (100%) injected heroin, and HIV prevalence was 30%. 144 of the 603 participants in the 2014 survey and 152 of the 600 PWID who had received lighters were "recaptured" in the 2016 survey. After adjusting for police suppression of drug use hotspots and conducting sensitivity analyses, our best estimate of the population size from the lighter recapture was 4617 (95% CI: 4090-5143), and our best estimate from the 2014 survey recapture was 5220 (95% CI: 4568-5872). A combined best estimate of the PWID population in Hai Phong is 5000, range 4000-6000. CONCLUSIONS: The capture/recapture studies produced consistent estimates. Adding a lighter/token distribution to planned RDS surveys may provide an inexpensive method for estimating PWID population size. Analyses of the estimates should include contextual information about the local drug scene.

Convergence of HIV seroprevalence among injecting and non-injecting drug users in New York City.

OBJECTIVE: To compare HIV prevalence among injecting and non-injecting heroin and cocaine users in New York City. As HIV is efficiently transmitted through the sharing of drug-injecting equipment, HIV infection has historically been higher among injecting drug users. DESIGN: Two separate cross-sectional surveys, both with HIV counseling and testing and drug use and HIV risk behavior questionnaires. METHODS: Injecting and non-injecting heroin and cocaine users recruited at detoxification and methadone maintenance treatment from 2001-2004 (n = 2121) and recruited through respondent-driven sampling from a research storefront in 2004 (n = 448). RESULTS: In both studies, HIV prevalence was nearly identical among current injectors (injected in the last 6 months) and heroin and cocaine users who had never injected: 13% [95% confidence interval (CI), 12-15%] among current injectors and 12% (95% CI, 9-16%) among never-injectors in the drug treatment program study, and 15% (95% CI, 11-19%) among current injectors and 17% (95% CI, 12-21%) among never injectors in the respondent driven sampling storefront study. The 95% CIs overlapped in all gender and race/ethnicity subgroup comparisons of HIV prevalence in both studies. CONCLUSIONS: The very large HIV epidemic among drug users in New York City appears to be entering a new phase, in which sexual transmission is of increasing importance. Additional prevention programs are needed to address this transition.

The transition from injection to non-injection drug use: long-term outcomes among heroin and cocaine users in New York City.

AIMS: To characterize heroin and cocaine users in New York City who have changed from injection to non-injection drug administration and to identify factors associated with long-term non-injection use. DESIGN: Two cross-sectional studies of heroin and cocaine users in New York City. SETTINGS AND PARTICIPANTS: New admissions were recruited at drug abuse treatment programs (2000-04) and respondent-driven sampling was used to recruit drug users from the community (2004). Both injecting and non-injecting drug users participated in each study. 'Former injectors' were defined operationally as people who had used heroin and/or cocaine in the 6 months prior to the interview and who had injected illicit drugs in the past, but whose most recent injection was more than 6 months before the study interview. 'Current' injectors were defined as people who had injected heroin and/or cocaine in the 6 months prior to the interview. MEASUREMENTS: A structured interview on drug use history was administered, and a serum sample was collected and tested for the human immunodeficiency virus (HIV). FINDINGS: A total of 104 former injectors was recruited for the drug abuse treatment program study, and 229 current injectors were recruited for the community recruitment study; 160 former injectors and 1731 current injectors were recruited from the drug abuse treatment study. Compared with the current injectors, former injectors were older and more likely to be African American. The former injectors reported long intervals since their most recent injection, a mean of 8 years in the drug abuse treatment program study and a mean of 12 years in the community recruitment study. The most common reasons for stopping injection drug use included concerns about health, social stigmatization and self-image, and preference for intranasal use as a route of drug administration. The results were highly consistent across the two studies. CONCLUSIONS: The transition from injection to non-injection use appears to be relatively stable behavior change for many former injectors, who report a decade or more without injecting. Developing a greater understanding of the transition from injection to stable non-injection drug use may provide insights into the natural histories of drug use and addiction.

Using Respondent-Driven Sampling to Recruit Illegal Drug Purchasers to Evaluate a Drug Market Intervention.

BACKGROUND: Violent drug markets are not as prominent as they once were in the United States, but they still exist and are associated with significant crime and lower quality of life. The drug market intervention (DMI) is an innovative strategy that uses focused deterrence, community engagement, and incapacitation to reduce crime and disorder associated with these markets. Although studies show that DMI can reduce crime and overt drug activity, one perspective is prominently missing from these evaluations: those who purchase drugs. OBJECTIVES: This study explores the use of respondent-driven sampling (RDS)-a statistical sampling method-to approximate a representative sample of drug users who purchased drugs in a targeted DMI market to gain insight into the effect of a DMI on market dynamics. METHODS: Using RDS, we recruited individuals who reported hard drug use (crack or powder cocaine, heroin, methamphetamine, or illicit use of prescriptions opioids) in the last month to participate in a survey. The main survey asked about drug use, drug purchasing, and drug market activity before and after DMI; a secondary survey asked about network characteristics and recruitment. CONCLUSIONS: Our sample of 212 respondents met key RDS assumptions, suggesting that the characteristics of our weighted sample approximate the characteristics of the drug user network. The weighted estimates for market purchasers are generally valid for inferences about the aggregate population of customers, but a larger sample size is needed to make stronger inferences about the effects of a DMI on drug market activity.

Design and baseline findings of a large-scale rapid response to an HIV outbreak in people who inject drugs in Athens, Greece: the ARISTOTLE programme.

AIMS: To (i) describe an intervention implemented in response to the HIV-1 outbreak among people who inject drugs (PWIDs) in Greece (ARISTOTLE programme), (ii) assess its success in identifying and testing this population and (iii) describe socio-demographic characteristics, risk behaviours and access to treatment/prevention, estimate HIV prevalence and identify risk factors, as assessed at the first participation of PWIDs. DESIGN: A 'seek, test, treat, retain' intervention employing five rounds of respondent-driven sampling. SETTING: Athens, Greece (2012-13). PARTICIPANTS: A total of 3320 individuals who had injected drugs in the past 12 months. INTERVENTION: ARISTOTLE is an intervention that involves reaching out to high-risk, hard-to-reach PWIDs ('seek'), engaging them in HIV testing and providing information and materials to prevent HIV ('test') and initiating and maintaining anti-retroviral and opioid substitution treatment for those testing positive ('treat' and 'retain'). MEASUREMENTS: Blood samples were collected for HIV testing and personal interviews were conducted. FINDINGS: ARISTOTLE recruited 3320 PWIDs during the course of 13.5 months. More than half (54%) participated in multiple rounds, resulting in 7113 visits. HIV prevalence was 15.1%. At their first contact with the programme, 12.5% were on opioid substitution treatment programmes and the median number of free syringes they had received in the preceding month was 0. In the multivariable analysis, apart from injection-related variables, homelessness was a risk factor for HIV infection in male PWIDs [odds ratio (OR) yes versus no = 1.89, 95% confidence interval (CI) = 1.41, 2.52] while, in female PWIDS, the number of sexual partners (OR for > 5 versus one partner in the past year = 4.12, 95% CI = 1.93, 8.77) and history of imprisonment (OR yes versus no = 2.76, 95% CI = 1.43, 5.31) were associated with HIV. CONCLUSIONS: In Athens, Greece, the ARISTOTLE intervention for identifying HIV-positive people among people who inject drugs (PWID) facilitated rapid identification of a hidden population experiencing an outbreak and provided HIV testing, counselling and linkage to care. According to ARISTOTLE data, the 2011 HIV outbreak in Athens resulted in 15% HIV infection among PWID. Risk factors for HIV among PWID included homelessness in men and history of imprisonment and number of sexual partners in women.

Increasing drug users' adherence to HIV treatment: results of a peer-driven intervention feasibility study.

Active drug users with HIV infection suffer from both low utilization of, and adherence to, primary care. Combining drug treatment and primary care on-site reduces these problems significantly because it creates a social support structure; treatment program staffs can monitor patients' adherence and provide ongoing encouragement. But in the United States, only a very small minority of HIV + drug users receive this demonstrably effective form of care. We report the results from a feasibility study of an alternative support structure, termed a "peer-driven intervention", that serves as a functional equivalent to drug treatment for increasing drug users' adherence to HIV therapeutics. The six-month study included 14 adult active drug users receiving medical care for HIV disease in New Haven, Connecticut. As a health advocate, each subject was assigned and asked to meet with another subject once a week at the project's storefront to provide peer support and counseling. As a peer, each subject was assigned and asked to meet with another health advocate once a week to receive support in keeping up his or her medical care. No two subjects played both roles for one another. Advocates earned nominal monetary rewards for eliciting positive responses from their peers in keeping clinical appointments, responding to physicians' referrals, picking up prescriptions on time and attending weekly meetings with the advocate. The results of the study suggest that an alternative social support structure to drug treatment is feasible for increasing active drug users' adherence to medical care. Innovative mechanisms that harness drug users' peer pressure to promote positive behavioral changes deserve greater study.

Assessing the assumptions of respondent-driven sampling in the national HIV Behavioral Surveillance System among injecting drug users.

Several assumptions determine whether respondent-driven sampling (RDS) is an appropriate sampling method to use with a particular group, including the population being recruited must know one another as members of the group (i.e., injection drug users [IDUs] must know each other as IDUs) and be networked and that the sample size is small relative to the overall size of the group. To assess these three assumptions, we analyzed city-specific data collected using RDS through the US National HIV Behavioral Surveillance System among IDUs in 23 cities. Overall, 5% of non-seed participants reported that their recruiter was "a stranger." 20 cities with multiple field sites had ≥1 cross-recruitment, a proxy for linked networks. Sample sizes were small in relation to the IDU population size (median = 2.3%; range: 0.6%- 8.0%). Researchers must evaluate whether these three assumptions were met to justify the basis for using RDS to sample specific populations.

Ethical and regulatory considerations in HIV prevention studies employing respondent-driven sampling.

OBJECTIVE: To review the scientific, ethical, and regulatory literature because no official guidance exists on remuneration for participant-driven recruitment or on investigators' responsibilities for informing participants of their discordant partnerships. METHODS: We reviewed the studies that used RDS to recruit injection-drug users (IDUs), 1995-2006, and the relevant scientific, ethical, and regulatory literature that shed light on arguments for and against practices that can be used in studies employing RDS. RESULTS: Concerns that payments can be used to buy drugs or can subvert altruistic motivations for study participation are not supported by the literature. Concerns about peers' coercion to barter coupons or surrender payments are offset by safeguards used in RDS. Remuneration shows respect for participants' time and effort in recruitment and judgment to use remuneration for personal needs. Tension between ensuring participants' confidentiality and protecting the health of network members highlights difference in roles between investigators and health care providers. Investigators can choose to rely on public health agencies for partner notification services, ask participants if and how they would like to be informed of their discordant partnerships, and offer training on how to disclose HIV status and adopt risk-reduction and harm-reduction behaviours. CONCLUSION: Clarifying ethical and regulatory considerations is important for research sponsors, institutional review boards (IRBs), ethics review committees (ERCs), investigators, and participants. We provide a checklist of ethics and regulatory variables to be included as feasible in future studies to enhance development of evidence-based ethics.

Parsing social network survey data from hidden populations using stochastic context-free grammars.

BACKGROUND: Human populations are structured by social networks, in which individuals tend to form relationships based on shared attributes. Certain attributes that are ambiguous, stigmatized or illegal can create a OhiddenO population, so-called because its members are difficult to identify. Many hidden populations are also at an elevated risk of exposure to infectious diseases. Consequently, public health agencies are presently adopting modern survey techniques that traverse social networks in hidden populations by soliciting individuals to recruit their peers, e.g., respondent-driven sampling (RDS). The concomitant accumulation of network-based epidemiological data, however, is rapidly outpacing the development of computational methods for analysis. Moreover, current analytical models rely on unrealistic assumptions, e.g., that the traversal of social networks can be modeled by a Markov chain rather than a branching process. METHODOLOGY/PRINCIPAL FINDINGS: Here, we develop a new methodology based on stochastic context-free grammars (SCFGs), which are well-suited to modeling tree-like structure of the RDS recruitment process. We apply this methodology to an RDS case study of injection drug users (IDUs) in Tijuana, México, a hidden population at high risk of blood-borne and sexually-transmitted infections (i.e., HIV, hepatitis C virus, syphilis). Survey data were encoded as text strings that were parsed using our custom implementation of the inside-outside algorithm in a publicly-available software package (HyPhy), which uses either expectation maximization or direct optimization methods and permits constraints on model parameters for hypothesis testing. We identified significant latent variability in the recruitment process that violates assumptions of Markov chain-based methods for RDS analysis: firstly, IDUs tended to emulate the recruitment behavior of their own recruiter; and secondly, the recruitment of like peers (homophily) was dependent on the number of recruits. CONCLUSIONS: SCFGs provide a rich probabilistic language that can articulate complex latent structure in survey data derived from the traversal of social networks. Such structure that has no representation in Markov chain-based models can interfere with the estimation of the composition of hidden populations if left unaccounted for, raising critical implications for the prevention and control of infectious disease epidemics.

High-risk sexual and drug using behaviors among male injection drug users who have sex with men in 2 Mexico-US border cities.

OBJECTIVES: The population of Latino men who have sex with men (MSM) and who are also injection drug users (IDUs) is understudied. We explored risk behaviors of MSM/IDUs compared with other male IDUs in 2 Mexican border cities. STUDY DESIGN: In 2005, IDUs who had injected within the previous 30 days were recruited using respondent-driven sampling (RDS) in Tijuana and Ciudad Juárez. They underwent antibody testing for HIV, HCV, and syphilis and interviewer-administered surveys. Men were categorized as MSM if they reported > or =1 lifetime male partners. Logistic regression was used to compare MSM/IDUs with non-MSM/IDUs. RESULTS: A third (31%) of 377 male IDUs were categorized as MSM (47% in Tijuana and 13% in Ciudad Juárez, P <0.01). Combined RDS-adjusted prevalence of HIV and Hepatitis C was 3% (95% CI: 1, 5) and 96%, (95% CI: 94, 99) respectively, while 17% (95% CI: 2, 36) of MSM and 8% (95% CI: 3, 12) of non-MSM tested positive for syphilis antibody. In multivariate logistic regression adjusted for site, MSM/IDUs were more likely than non-MSM/IDUs to have ever used inhalants (OR: 3.4; 95% CI: 1.8, 6.2) or oral tranquilizers (OR: 2.4; 95% CI: 1.3, 4.6), received treatment for a drug problem (OR:1.9; 95% CI: 1.1, 3.2) shared needles in the last six months (OR: 2.1; 95% CI: 1.0, 4.2) and also had higher numbers of lifetime female partners (log-transformed continuous variable, OR: 1.6; 95% CI: 1.2, 2.1). CONCLUSIONS: In these Mexican cities, the proportion of MSM among male IDUs was high. Compared with other male IDUs, MSM/IDUs were more likely to engage in behaviors placing them at risk of acquiring HIV/STIs. Culturally appropriate interventions targeting Latino MSM/IDUs are warranted.

Effectiveness of respondent-driven sampling for recruiting drug users in New York City: findings from a pilot study.

A number of sampling methods are available to recruit drug users and collect HIV risk behavior data. Respondent-driven sampling (RDS) is a modified form of chain-referral sampling with a mathematical system for weighting the sample to compensate for its not having been drawn randomly. It is predicated on the recognition that peers are better able than outreach workers and researchers to locate and recruit other members of a "hidden" population. RDS provides a means of evaluating the reliability of the data obtained and also allows inferences about the characteristics of the population from which the sample is drawn. In this paper we present findings from a pilot study conducted to assess the effectiveness of RDS to recruit a large and diversified group of drug users in New York City. Beginning with eight seeds (i.e., initial recruits) we recruited 618 drug users (injecting and non-injecting) in 13 weeks. The data document both cross-gender and cross-race and -ethnic recruitment as well as recruitment across drug-use status. Sample characteristics are similar to the characteristics of the drug users recruited in other studies conducted in New York City. The findings indicate that RDS is an effective sampling method for recruiting diversified drug users to participate in HIV-related behavioral surveys.

From networks to populations: the development and application of respondent-driven sampling among IDUs and Latino gay men.

One of the challenges in studying HIV-risk behaviors among gay men is gathering information from a non-biased sample, as traditional probability sampling methods cannot be applied in gay populations. Respondent-Driven Sampling (RDS) has been proposed as a reliable and bias-free method to recruit "hidden" populations, such as gay men. The aim of this study is to assess the feasibility and effectiveness of RDS to sample Latino gay men and transgender persons. This was carried out when we used RDS to recruit participants into a study that investigated community involvement on HIV/AIDS sexual risk behaviors among Latino gay and bisexual men, and transgender (male-to-female) persons in Chicago and San Francisco. The population coverage of RDS was then compared to simulated time-location sampling (TLS). Recruitment differences were observed across cities, but the samples were comparable. RDS showed broader population coverage than TLS, especially among individuals at high risk for HIV.