Respondent-driven Sampling Recruitment Research Articles

Using Respondent-Driven Sampling (RDS) to Identify the Healthcare Needs among Women of Reproductive Age Who Migrated from Venezuela to Brazil, 2018-2021.

In 2021, an RDS survey was conducted among Venezuelan migrant women of reproductive age who migrated to two Brazilian cities (Manaus and Boa Vista) from 2018 to 2021. To start the RDS recruitment, we chose seeds non-randomly in both cities. The study variables were age, educational level, self-rated health, pregnancy, migratory status and use of health services. We estimated the prevalence, confidence intervals and homophily effects by variable category. We used a multivariate logistic regression model to identify the main factors associated with healthcare use. A total of 761 women were recruited in Manaus and 1268 in Boa Vista. Manaus showed more irregular migrants than Boa Vista. The main reasons for using health services were as follows: illness, disease prevention and prenatal care. The logistic regression model showed the use of health services was associated with educational level and healthcare needs but not with migratory status. The social inclusion of Venezuelan migrants is extremely relevant, although many challenges must be overcome. The strategy of the Brazilian Federal Government for providing humanitarian assistance to Venezuelan migrants should be expanded to include and facilitate their integration into labor markets, access to healthcare and education, benefiting both migrants and the Brazilian people by reducing social inequality.

A comprehensive assessment of quality management methods in the SMESH study

BackgroundThis paper aims to instigate discussion and publication of methodologies applied to enhance quality management through comprehensive scientific reports. It provides a detailed description of the design, implementation, and results of the quality control program employed in the SMESH study.MethodsCross-sectional, multicenter, national study designed to assess the prevalence of human papillomavirus in sex workers and in men who have sex with men (MSM). Respondent-driven sampling recruitment was used. An online system was developed for the study and checkpoints were defined for data entry. The system checked the quality of biological samples and performed a retest with part of the sample.ResultsA total of 1.598 participants (442 sex workers and 1.156 MSM) were included. Fifty-four health professionals were trained for face-to-face data collection. The retest showed Kappa values ranging between 0.3030 and 0.7663.ConclusionThe retest data were mostly classified as indicating a strong association. The data generated by the checkpoints showed the successful implementation of the quality control program.

Recruiting hard-to-reach populations via respondent driven sampling for mobile phone surveys in Colombia: a qualitative study

ABSTRACT Background Uptake of mobile phone surveys (MPS) is increasing in many low- and middle-income countries, particularly within the context of data collection on non-communicable diseases (NCDs) behavioural risk factors. One barrier to collecting representative data through MPS is capturing data from older participants. Respondent driven sampling (RDS) consists of chain-referral strategies where existing study subjects recruit follow-up participants purposively based on predefined eligibility criteria. Adapting RDS strategies to MPS efforts could, theoretically, yield higher rates of participation for that age group. Objective To investigate factors that influence the perceived acceptability of a RDS recruitment method for MPS involving people over 45 years of age living in Colombia. Methods An MPS recruitment strategy deploying RDS techniques was piloted to increase participation of older populations. We conducted a qualitative study that drew from surveys with open and closed-ended items, semi-structured interviews for feedback, and focus group discussions to explore perceptions of the strategy and barriers to its application amongst MPS participants. Results The strategy’s success is affected by factors such as cultural adaptation, institutional credibility and public trust, data protection, and challenges with mobile phone technology. These factors are relevant to individuals’ willingness to facilitate RDS efforts targeting hard-to-reach people. Recruitment strategies are valuable in part because hard-to-reach populations are often most accessible through their contacts within their social network who can serve as trust liaisons and drive engagement. Conclusions These findings may inform future studies where similar interventions are being considered to improve access to mobile phone-based data collection amongst hard-to-reach groups.

High Seroprevalence of Hepatitis C Virus Among Cisgender Women Who Exchange Sex in the Seattle, Washington Area.

Injection drug use (IDU) is a risk factor for hepatitis C virus (HCV) acquisition and occurs disproportionately among women who exchange sex (WES). However, little is known about HCV epidemiology in this population. We estimated HCV seroprevalence, identified correlates of HCV seropositivity, and characterized social networks by HCV serostatus and IDU history among WES in the Seattle, Washington, area. This was a secondary analysis of data from the 2016 National HIV Behavioral Surveillance survey in the Seattle, Washington area, a cross-sectional survey that used respondent-driven sampling (RDS) to enroll WES for money or drugs (N = 291). All participants were offered rapid HCV-antibody testing. We estimated HCV seropositivity and used log regression methods to estimate crude and adjusted prevalence ratios (PRs) for correlates of HCV seropositivity among WES. Using RDS recruitment chain data, we computed homophily indices to estimate the extent to which participants were likely to recruit another participant with the same HCV serostatus and IDU history. In the study sample of WES in the Seattle, Washington area, 79% reported lifetime IDU and 60% were HCV seropositive. HCV seropositivity was strongly associated with ever injecting drugs (PRadj: 7.7 [3.3, 18.0]). The RDS homophily scores for HCV seropositivity (0.07) and ever injecting drugs (0.02) suggested that participants did not tend to recruit others with the same characteristics beyond what would be expected by chance. Among this sample of WES in Seattle, Washington area, HCV seroprevalence was high and strongly associated with a history of IDU. The high burden of HCV among WES suggests this marginalized group would benefit from additional harm reduction services and targeted HCV treatment campaigns to reduce forward transmission. We saw little evidence of preferential recruitment among WES who were HCV seropositive or reported a history of IDU, suggesting the potential futility of peer-based referrals for HCV treatment.

Examining differential success in recruitment using respondent driven sampling (RDS) in a multi-site study of gay, bisexual and other men who have sex with men

Background The Engage Study is a longitudinal biobehavioral cohort study of gay, bisexual and other men who have sex with men (GBM) in Toronto, Montreal, and Vancouver. Baseline data (2,449 participants) were collected from February 2017 - August 2019 using respondent-driven sampling (RDS). Recruitment in Montreal required fewer seeds, had a much shorter recruitment period, and recruited the largest sample.MethodsTo better understand why RDS recruitment was more successful in Montreal compared to other sites, we conducted an analysis to examine RDS recruitment characteristics for GBM in each of the three study sites, explore demographic characteristics and measures of homophily, that is, the tendency of individuals to recruit other study participants who are like themselves, and compared motivations for study participation.ResultsMontreal had the greatest proportion of participants over the age of 45 (29.1% in Montreal, 24.6% in Vancouver, and 21.0% in Toronto) and the highest homophily for this age group, but homophily was high across the three cities. Montreal also reported the lowest percentage of participants with an annual income greater or equal to $60,000 (7.9% in Montreal, 13.1% in Vancouver and 10.6% in Toronto), but homophily was similar across all three cities. The majority of participants indicated interest in sexual health and HIV as the main reason for participating (36.1% in Montreal, 34.7% in Vancouver, and 29.8% in Toronto). Financial interest as the main reason for participation was low (12.7% in Montreal, 10.6% in Vancouver, and 5.7% in Toronto).ConclusionTaken together, although we found some differences in study demographic characteristics and homophily scores, we were unable to fully explain the different recruitment success based on the data available. Our study underlines the fact that success of RDS implementation may vary by unknown factors, and that researchers should be proactive and flexible to account for variability.

Employing Respondent Driven Sampling (RDS) to recruit people who inject drugs (PWID) and other hard-to-reach populations during COVID-19: Lessons learned.

BackgroundRespondent Driven Sampling (RDS) is an effective sampling strategy to recruit hard-to-reach populations but the impact of the COVID-19 pandemic on the use of this strategy in the collection of data involving human subjects, particularly among marginalized and vulnerable populations, is not known. Based on an ongoing study using RDS to recruit and study the interactions between HIV infection, injection drug use, and the microbiome in Puerto Rico, this paper explores the effectiveness of RDS during the pandemic and provided potential strategies that could improve recruitment and data collection.ResultsRDS was employed to evaluate its effectiveness in recruiting a group of people who inject drugs (PWID) and controls (N = 127) into a study in the midst of the COVID-19 pandemic. The participants were distributed among three subsets: 15 were HIV+ and PWID, 58 were HIV- PWID, and 54 were HIV+ and not PWID.FindingsResults show that recruitment through peer networks using RDS was possible across all sub-groups. Yet, while those in the HIV+ PWID sub-group managed to recruit from other-sub groups of HIV- PWID and HIV+, this occurred at a lower frequency.ConclusionDespite the barriers introduced by COVID-19, it is clear that even in this environment, RDS continues to play a powerful role in recruiting hard-to-reach populations. Yet, more attention should be paid at how future pandemics, natural disasters, and other big events might affect RDS recruitment of vulnerable and hard-to-reach populations.

Using large-scale respondent driven sampling to monitor the end of an HIV epidemic among persons who inject drugs in Hai Phong, Viet Nam.

AimsTo describe the use of large-scale respondent driven sampling (RDS) surveys to demonstrate the “end of an HIV epidemic” (HIV incidence < 0.5/100 person-years) among persons who inject drugs (PWID) in a middle-income country. Large sample sizes are needed to convincingly demonstrate very low incidence rates.Methods4 large surveys (Ns approximately 1500 each) were conducted among PWID in Hai Phong, Vietnam in 2016–2019. Respondent driven sampling (RDS) with a modification to add snowball sampling was used for recruiting participants. HIV incidence was measured through recency testing, repeat participants across multiple surveys and in a cohort study of PWID recruited from the surveys. RDS analytics (time to equilibria and homophilies for major variables) were used to assess similarities/differences in RDS only versus RDS plus snowball recruiting. Characteristics were compared among respondents recruited through standard RDS recruitment versus through snowball sampling. An overall assessment of the robustness of RDS to modification was made when adding a snowball sampling recruitment.ResultsRDS recruiting was very efficient in the first 5 weeks of each survey with approximately 180 respondents recruited per week. Recruiting then slowed considerably, and snowball sampling (permitting an individual respondent to recruit large numbers of new respondents) was added to the existing RDS recruiting. This led to recruiting within 13–14 weeks of 1383, 1451, 1444 and 1268 respondents, close to the target of 1500 respondents/survey.Comparisons of participants recruited through standard RDS method and respondents recruited through snowball methods showed very few significant differences. RDS analytics (quickly reaching equilibria, low homophilies) were favorable for both RDS recruited and total numbers of participants in each survey. DRug use and Infections in ViEtnam (DRIVE) methods have now been officially adopted in other provinces.ConclusionsRDS appears to be quite robust with respect to adding a modest number of participants recruited through snowball sampling. Large sample sizes can provide compelling evidence for “ending an HIV epidemic” to policy makers in a PWID population in a middle income country setting.

Comparing recruitment strategies to engage hard-to-reach men who have sex with men living with HIV with unsuppressed viral loads in four US cities: Results from HPTN 078.

IntroductionThere is an urgent need to identify men who have sex with men (MSM) living with HIV with unsuppressed viral loads to prevent transmission. Though respondent‐driven sampling (RDS) is traditionally used for hard‐to‐reach populations, we compare how RDS and direct recruitment (DR) perform in identifying MSM living with HIV with unsuppressed viral loads and identifying MSM with socio‐demographics characteristic of hard‐to‐reach populations.MethodsThis is a cross‐sectional analysis among 1305 MSM who were recruited from March 2016 to December 2017 for a case management intervention trial (HPTN 078). We recruited participants across four cities using RDS and DR methods: Birmingham, AL; Atlanta, GA; Baltimore, MD; and Boston, MA. Participants completed a socio‐demographic questionnaire and underwent HIV testing. We compare the proportion of MSM with HIV and unsuppressed viral loads (HIV RNA ≥ 1000 copies/ml) based on recruitment method using Pearson chi‐square tests. We also compare differences in race, income, healthcare coverage, education, sexual orientation, hidden sexuality and comfort with participating in the LGBT community between recruitment methods and perform non‐parametric trend tests to see how demographics change across RDS recruitment waves.ResultsRDS recruited 721 men (55.2%) and DR yielded 584 men (44.8%). Overall, 69% were living with HIV, of whom 18% were not virally suppressed. HIV prevalence was higher among those recruited via DR (84%) compared to RDS (58%), p < 0.0001. Twenty per cent of DR recruits were not virally suppressed compared to 15% of RDS, though this was not significant. DR yielded a significantly higher proportion of Black participants and those with less than a high school diploma. The prevalence of low income, no healthcare coverage, bisexuality and hidden sexuality increased across RDS waves.ConclusionsDR was more efficient in identifying MSM living with HIV with unsuppressed viral loads; however, there was a higher proportion of hard‐to‐reach MSM who were low income, lacked health coverage, were bisexual and were not open with their sexuality in deeper waves of RDS. Researchers should consider supplementing RDS recruitment with DR efforts if aiming to identify MSM with unsuppressed viral loads via RDS.

Using Social Networks to Supplement RDD Telephone Surveys to Oversample Hard-to-Reach Populations: A New RDD+RDS Approach

Random digit dialing (RDD) telephone sampling, although experiencing declining response rates, remains one of the most accurate and cost-effective data collection methods for generating national population-based estimates. Such methods, however, are inefficient when sampling hard-to-reach populations because the costs of recruiting sufficient sample sizes to produce reliable estimates tend to be cost prohibitive. The authors implemented a novel respondent-driven sampling (RDS) approach to oversample cigarette smokers and lesbian, gay, bisexual, and transgender (LGBT) people. The new methodology selects RDS referrals or seeds from a probability-based RDD sampling frame and treats the social networks as clusters in the weighting and analysis, thus eliminating the intricate assumptions of RDS. The authors refer to this approach as RDD+RDS. In 2016 and 2017, a telephone survey was conducted on tobacco-related topics with a national sample of 4,208 U.S. adults, as well as 756 referral-based respondents. The RDD+RDS estimates were comparable with stand-alone RDD estimates, suggesting that the addition of RDS responses from social networks improved the precision of the estimates without introducing significant bias. The authors also conducted an experiment to determine whether the number of recruits would vary on the basis of how the RDS recruitment question specified the recruitment population (closeness of relationship, time since last contact, and LGBT vs. tobacco user), and significant differences were found in the number of referrals provided on the basis of question wording. The RDD+RDS sampling approach, as an adaptation of standard RDD methodology, is a practical tool for survey methodologists that provides an efficient strategy for oversampling rare or elusive populations.

Associating conditional cash transfer to universal access to treatment could be the solution to the HCV epidemic among drug users (DUs)

BackgroundTo understand the limits of HCV screening programs to reach all drug users (DUs).MethodThe association of the recruitment of a representative sample of a population of DUs in a specific area with the use of a questionnaire that included 250 items allowed the use of uni- and multifactorial analysis to explore the relationship between HCV screening and dimensions until now restricted to qualitative studies.ResultsWe recruited, in less than 2 months, 327 DUs representing about 6% of the total population of DUs. They belonged to a single community whose drug use was the only common characteristic. While almost all DUs (92.6%) who had access to care providers had been screened, this proportion was much lower in out-of-care settings (64%). HCV prevalence among those who had performed a test was low (22.8%). For DUs, the life experience of hepatitis C has not changed in the last 10 years. Screening, studied for the first time according to this life experience, was not influenced by a rational knowledge of the risk taken or the knowledge of treatment efficacy, showing a gap between DUs’ representations and medical recommendations which explains the low level of active screening. Police crackdown on injections, disrupting the previous illusion of safe practices, was the only prior history leading to active screenings. Screenings were related to an access to care providers. GPs held a preponderant position as a source of information and care by being able to give appropriate answers regarding hepatitis C and prescribing opioid substitution treatments (OST). If 48 % of DUs screened positive for HCV had been treated, half of them had been prescribed before 2006.ConclusionWhile hepatitis has become a major issue for society and, consequently, for services for DUs (SDUs) and GPs, it is not the case for DUs. A widespread screening, even in a city where the offer of care is diversified and free, seems unlikely to reach a universal HCV screening over a short time. The model of respondent-driven sampling recruitment could be a new approach to conditional cash transfer, recruiting and treating DUs who remain outside the reach of care providers, a prerequisite for the universal access to HCV treatments to impact the HCV epidemic.

Hidden Population Size Estimation From Respondent-Driven Sampling: A Network Approach

ABSTRACTEstimating the size of stigmatized, hidden, or hard-to-reach populations is a major problem in epidemiology, demography, and public health research. Capture–recapture and multiplier methods are standard tools for inference of hidden population sizes, but they require random sampling of target population members, which is rarely possible. Respondent-driven sampling (RDS) is a survey method for hidden populations that relies on social link tracing. The RDS recruitment process is designed to spread through the social network connecting members of the target population. In this article, we show how to use network data revealed by RDS to estimate hidden population size. The key insight is that the recruitment chain, timing of recruitments, and network degrees of recruited subjects provide information about the number of individuals belonging to the target population who are not yet in the sample. We use a computationally efficient Bayesian method to integrate over the missing edges in the subgraph of recruited individuals. We validate the method using simulated data and apply the technique to estimate the number of people who inject drugs in St. Petersburg, Russia. Supplementary materials for this article are available online.

Total Survey Error and Respondent Driven Sampling: Focus on Nonresponse and Measurement Errors in the Recruitment Process and the Network Size Reports and Implications for Inferences

Abstract This study attempted to integrate key assumptions in Respondent-Driven Sampling (RDS) into the Total Survey Error (TSE) perspectives and examine TSE as a new framework for a systematic assessment of RDS errors. Using two publicly available data sets on HIV-at-risk persons, nonresponse error in the RDS recruitment process and measurement error in network size reports were examined. On nonresponse, the ascertained partial nonresponse rate was high, and a substantial proportion of recruitment chains died early. Moreover, nonresponse occurred systematically: recruiters with lower income and higher health risks generated more recruits; and peers of closer relationships were more likely to accept recruitment coupons. This suggests a lack of randomness in the recruitment process, also shown through sizable intra-chain correlation. Self-reported network sizes suggested measurement error, given their wide dispersion and unreasonable reports. This measurement error has further implications for the current RDS estimators, which use network sizes as an adjustment factor on the assumption of a positive relationship between network sizes and selection probabilities in recruitment. The adjustment resulted in nontrivial unequal weighting effects and changed estimates in directions that were difficult to explain and, at times, illogical. Moreover, recruiters’ network size played no role in actual recruitment. TSE may serve as a tool for evaluating errors in RDS, which further informs study design decisions and inference approaches.

THE GRAPHICAL STRUCTURE OF RESPONDENT-DRIVEN SAMPLING.

Respondent-driven sampling (RDS) is a chain-referral method for sampling members of hidden or hard-to-reach populations, such as sex workers, homeless people, or drug users, via their social networks. Most methodological work on RDS has focused on inference of population means under the assumption that subjects' network degree determines their probability of being sampled. Criticism of existing estimators is usually focused on missing data: the underlying network is only partially observed, so it is difficult to determine correct sampling probabilities. In this article, the author shows that data collected in ordinary RDS studies contain information about the structure of the respondents' social network. The author constructs a continuous-time model of RDS recruitment that incorporates the time series of recruitment events, the pattern of coupon use, and the network degrees of sampled subjects. Together, the observed data and the recruitment model place a well-defined probability distribution on the recruitment-induced subgraph of respondents. The author shows that this distribution can be interpreted as an exponential random graph model and develops a computationally efficient method for estimating the hidden graph. The author validates the method using simulated data and applies the technique to an RDS study of injection drug users in St. Petersburg, Russia.

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

ABSTRACTCombined 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.

Lessons learned from respondent-driven sampling recruitment in Nairobi: experiences from the field.

BackgroundRespondent-driven sampling (RDS) is used in a variety of settings to study hard-to-reach populations at risk for HIV and sexually transmitted infections. However, practices leading to successful recruitment among diverse populations in low-resource settings are seldom reported. We implemented the first, integrated, bio-behavioural surveillance survey among men who have sex with men, female sex workers and people who injected drugs in Nairobi, Kenya.MethodsThe survey period was June 2010 to March 2011, with a target sample size of 600 participants per key populations. Formative research was initially conducted to assess feasibility of the survey. Weekly monitoring reports of respondent characteristics and recruitment chain graphs from NetDraw illustrated patterns and helped to fill recruitment gaps.ResultsRDS worked well with men who have sex with men and female sex workers with recruitment initiating at a desirable pace that was maintained throughout the survey. Networks of people who injected drugs were well-integrated, but recruitment was slower than the men who have sex with men and female sex workers surveys.ConclusionBy closely monitoring RDS implementation and conducting formative research, RDS studies can effectively develop and adapt strategies to improve recruitment and improve adherence to the underlying RDS theory and assumptions.

High prevalence of HIV, chlamydia and gonorrhoea among men who have sex with men and transgender women attending trusted community centres in Abuja and Lagos, Nigeria.

IntroductionSexually transmitted infection (STI) and HIV prevalence have been reported to be higher amongst men who have sex with men (MSM) in Nigeria than in the general population. The objective of this study was to characterize the prevalence of HIV, chlamydia and gonorrhoea in this population using laboratory-based universal testing.MethodsTRUST/RV368 represents a cohort of MSM and transgender women (TGW) recruited at trusted community centres in Abuja and Lagos, Nigeria, using respondent-driven sampling (RDS). Participants undergo a structured comprehensive assessment of HIV-related risks and screening for anorectal and urogenital Chlamydia trachomatis and Neisseria gonorrhoeae, and HIV. Crude and RDS-weighted prevalence estimates with 95% confidence intervals (CIs) were calculated. Log-binomial regression was used to explore factors associated with prevalent HIV infection and STIs.ResultsFrom March 2013 to January 2016, 862 MSM and TGW (316 in Lagos and 546 in Abuja) underwent screening for HIV, chlamydia and gonorrhoea at study enrolment. Participants’ median age was 24 years [interquartile range (IQR) 21–27]. One-third (34.2%) were identified as gay/homosexual and 65.2% as bisexual. The overall prevalence of HIV was 54.9%. After adjusting for the RDS recruitment method, HIV prevalence in Abuja was 43.5% (95% CI 37.3–49.6%) and in Lagos was 65.6% (95% CI 54.7–76.5%). The RDS-weighted prevalence of chlamydia was 17.0% (95% CI 11.8–22.3%) in Abuja and 18.3% (95% CI 11.1–25.4%) in Lagos. Chlamydia infection was detected only at the anorectal site in 70.2% of cases. The RDS-weighted prevalence of gonorrhoea was 19.1% (95% CI 14.6–23.5%) in Abuja and 25.8% (95% CI 17.1–34.6%) in Lagos. Overall, 84.2% of gonorrhoea cases presented with anorectal infection only. Over 95% of STI cases were asymptomatic. In a multivariable model, increased risk for chlamydia/gonorrhoea was associated with younger age, gay/homosexual sexual orientation and higher number of partners for receptive anal sex. HIV infection was associated with older age, female gender identity and number of partners for receptive anal sex.ConclusionsThere is a high burden of infection with HIV and asymptomatic chlamydia and gonorrhoea among MSM and TGW in Nigeria. Most cases would have been missed without anorectal screening. Interventions are needed to target this population for appropriate STI screening and management beginning at a young age.

Adapting chain referral methods to sample new migrants

Background: Demographic research on migration requires representative samples of migrant populations. Yet recent immigrants, who are particularly informative about current migrant flows, are difficult to capture even in specialist surveys. Respondent-driven sampling (RDS), a chain referral sampling and analysis technique, potentially offers the opportunity to achieve population-level inference of recently arrived migrant populations. Objective: We evaluate the attempt to use RDS to sample two groups of migrants, from Pakistan and Poland, who had arrived in the UK within the previous 18 months, and we present an alternative approach adapted to recent migrants. Methods: We discuss how connectedness, privacy, clustering, and motivation are expected to differ among recently arrived migrants, compared to typical applications of RDS. We develop a researcher-led chain referral approach, and compare success in recruitment and indicators of representativeness to standard RDS recruitment. Results: Our researcher-led approach led to higher rates of chain-referral, and enabled us to reach population members with smaller network sizes. The researcher-led approach resulted in similar recruiter-recruit transition probabilities to traditional RDS across many demographic and social characteristics. However, we did not succeed in building up long referral chains, largely due to the lack of connectedness of our target populations and some reluctance to refer. There were some differences between the two migrant groups, with less mobile and less hidden Pakistani men producing longer referral chains. Conclusions: Chain referral is difficult to implement for sampling newly arrived migrants. However, our researcher-led adaptation shows promise for less hidden and more stable recent immigrant populations. Contribution: The paper offers an evaluation of RDS for surveying recent immigrants and an adaptation that may be effective under certain conditions.

Hepatitis C virus seroprevalence among people who inject drugs and factors associated with infection in eight Russian cities.

BackgroundBehavioural surveillance among people who inject drugs (PWID) and testing for hepatitis C virus (HCV) and HIV is needed to understand the scope of both epidemics in at-risk populations and to suggest steps to improve their health.MethodsPWID were recruited using respondent-driven sampling (RDS) in eight Russian cities. A standardized survey was administered to collect sociodemographic and behavioral information. Blood specimens were obtained for serological testing for HCV and HIV-1. Data across the eight sites were pooled to identify individual-, network-, and city-level factors associated with positive HCV serostatus.ResultsAmong 2,596 PWID participating in the study, 1,837 tested positive for HCV (71%). The sample was 73% male and the mean age was 28. Very few PWID reported regular contact with harm reduction programs. Factors associated with testing positive for HCV were longer duration of injection drug use, testing positive for HIV-1, sharing non-syringe injection paraphernalia and water for rinsing syringes, and larger social network size. Factors negatively associated with HCV-positive serostatus were injecting with a used syringe and two city-level factors: longer mean RDS recruitment chain in a city and higher levels of injecting stimulants.ConclusionsHCV prevalence in all eight Russian cities is at the higher end of the range of HCV prevalence among PWID in Europe, which provides evidence that more resources, better prevention programs, and accelerated treatment targeting PWID are needed to control the HCV epidemic.

Comparison of Surveillance Sample Demographics Over Two Cycles of the National HIV Behavioral Surveillance Project, Houston, Texas

We examined differences in sample demographics across cycles of the National HIV Behavioral Surveillance project, that examines HIV risk behaviors among men who have sex with men (MSM), injection drug users (IDU), and heterosexuals living in areas of high HIV prevalence (HET). MSM were recruited through venue-based sampling, and IDU and HET through respondent driven sampling (RDS). RDS data were weighted to account for sampling bias. We compared crude prevalence estimates from MSM1 (2004) to those from MSM2 (2008) for demographic factors known to influence risky sexual and drug-use behaviors. We compared crude and adjusted prevalence estimates for IDU1 (2005) and IDU2 (2009) and HET1 (2006) and HET2 (2010). In the MSM cycle, we found differences in age, and the proportions seeking medical care and reporting a recent arrest. There were no differences in the comparison of crude and weighted estimates for the RDS collected samples, nor were there differences comparing HET1 and HET2 weighted estimates. IDU2 recruited a larger proportion of males, and had a higher percent who graduated from high school and who reported recent medical care and a previous HIV test. Differences across MSM cycles may be related to differences in venues identified for each cycle. Differences in the IDU cycles may be due to an effort on our part to increase the racial/ethnic and drug-use diversity of the sample in IDU2. Our findings show the importance of formative work for both venue-based and RDS samples to increase understanding of the dimensions that affect social networks and the dynamics of populations in space and time. With familiarity of the target population, we believe that both venue-based and RDS recruitment approaches for NHBS work well and can be used to evaluate changes in risky sexual and drug use behaviors and in HIV testing behaviors.

An Empirical Analysis of the Impact of Recruitment Patterns on RDS Estimates among a Socially Ordered Population of Female Sex Workers in China

Respondent-driven sampling (RDS) is a method for recruiting "hidden" populations through a network-based, chain and peer referral process. RDS recruits hidden populations more effectively than other sampling methods and promises to generate unbiased estimates of their characteristics. RDS's faithful representation of hidden populations relies on the validity of core assumptions regarding the unobserved referral process. With empirical recruitment data from an RDS study of female sex workers (FSWs) in Shanghai, we assess the RDS assumption that participants recruit nonpreferentially from among their network alters. We also present a bootstrap method for constructing the confidence intervals around RDS estimates. This approach uniquely incorporates real-world features of the population under study (e.g., the sample's observed branching structure). We then extend this approach to approximate the distribution of RDS estimates under various peer recruitment scenarios consistent with the data as a means to quantify the impact of recruitment bias and of rejection bias on the RDS estimates. We find that the hierarchical social organization of FSWs leads to recruitment biases by constraining RDS recruitment across social classes and introducing bias in the RDS estimates.