Evaluating Respondent-Driven Sampling in a Major Metropolitan Area: Comparing Injection Drug Users in the 2005 Seattle Area National HIV Behavioral Surveillance System Survey with Participants in the RAVEN and Kiwi Studies
Introduction
Injection drug users (IDU) are a population at elevated risk for several infections of public health importance, including HIV, hepatitis B, and hepatitis C 1, 2. Epidemiologic surveys of drug-injecting populations are important in measuring the prevalence of these diseases, evaluating public health measures to control them, identifying unmet needs and noting opportunities for prevention efforts. However, obtaining an unbiased sample of IDU has proved problematic because of the illegal nature of drug injection and the social marginalization of many IDU. Most methods of IDU recruitment contain well-recognized sources of bias, the effects of which are difficult or impossible to quantify 3, 4.
Respondent-driven sampling (RDS) has been proposed as an advantageous means of accessing hidden populations 5, 6, 7. In RDS, participants are given coupons with which to recruit their peers and offered payments when new recruits bring in the coupons. Theoretical reasoning and mathematical modeling propose that RDS methods can produce a study population from which unbiased estimates of the properties of a target population can be calculated. RDS offers further advantages in the ease of its implementation and the standardization of its methods, advantages which have made RDS attractive in international studies with limited resources 8, 9. While RDS is being used in a growing number of settings 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, there remains a need for empirical data evaluating how well RDS fulfills its promise in practice.
One means of evaluating RDS is to compare RDS results for a population with data on the same population obtained through other means (25). In 2005, IDU were surveyed for the National HIV Behavioral Surveillance system using RDS in 23 U.S. cities (the NHBS-IDU1 survey), including Seattle (25). Two previous studies, the Risk Activity Variables, Epidemiology and Network Study (RAVEN) and the Kiwi study, recruited Seattle area IDU using differing sampling strategies. Data on IDU diagnosed with HIV/AIDS during the period 2001–2005 were available through the HIV/AIDS Reporting System (HARS). We compared these different populations in terms of area of residence, age, race, gender, and drug most frequently injected, variables commonly used to characterize IDU populations 15, 18, 20, 26.
Section snippets
Recruitment
The methodology for surveying IDU in NHBS-IDU1 has been described (25). Participants were required to be 18 years of age or older; have injected in the previous 12 months; reside in King, Snohomish, or Island counties; and be able to communicate in English. In Seattle, 19 initial IDU (seeds) were each given three coupons to pass on to their injecting peers. Participants who completed a survey questionnaire were paid $20 and offered in turn coupons to distribute to their IDU peers. They received
Measures of RDS Recruitment
Of the 19 seeds interviewed, 10 recruited at least one study participant. Sixty percent of the sample population derived from one seed. Fifty-eight percent of participants recruited at least one new study participant; 31% of the coupons distributed were returned by new participants. The rate of recruitment was substantially slower than anticipated. At the fixed time the survey was terminated (as mandated by CDC protocol), data from 370 eligible participants were available for analysis; this was
Discussion
The NHBS-IDU1 results stand out from RAVEN, Kiwi, and HARS data in the high estimated proportion of participants from downtown Seattle residents and in the markedly older age distribution. Lacking a definitive gold standard, we cannot determine with assurance which of these populations, if any, accurately reflects the characteristics of Seattle area IDU. The various data sources in this report have strengths and weaknesses. RAVEN's random number–based sampling reduced volunteer bias, and the
References (44)
- et al.
Respondent-driven sampling to recruit MDMA users: a methodological assessment
Drug Alcohol Depend
(2005) - et al.
Respondent-driven sampling in the recruitment of illicit stimulant drug users in a rural setting: findings and technical issues
Addict Behav
(2007) - et al.
Respondent-driven sampling to assess characteristics and estimate the number of injection drug users in Bangkok, Thailand
Drug Alcohol Depend
(2007) Hepatitis Surveillance Report No. 61
(2006)- Centers for Disease Control and Prevention. HIV/AIDS Surveillance Report, 2005. Vol. 18. Available at:...
- et al.
Review of sampling hard-to-reach and hidden populations for HIV surveillance
AIDS
(2005) - et al.
Street and network sampling in evaluation studies of HIV risk-reduction interventions
AIDS Rev
(2002) Respondent-driven sampling: A new approach to the study of hidden populations
Social Problems
(1997)Respondent-driven sampling II: Deriving valid population estimates from chain-referral samples of hidden populations
Social Problems
(2002)- et al.
Sampling and estimation in hidden populations using respondent-driven sampling
Sociological Methodology
(2004)
Implementation challenges to using Respondent-Driven Sampling methodology for HIV biological and behavioral surveillance: field experiences in international settings
AIDS Behav
Using Respondent-Driven Sampling methodology for HIV biological and behavioral surveillance in international settings: a systematic review
AIDS Behav
Assessment of respondent driven sampling for recruiting female sex workers in two Vietnamese cities: reaching the unseen sex worker
J Urban Health
The effectiveness of respondent driven sampling for recruiting males who have sex with males in Dhaka, Bangladesh
AIDS Behav
An empirical comparison of respondent-driven sampling, time location sampling, and snowball sampling for behavioral surveillance in men who have sex with men, Fortaleza, Brazil
AIDS Behav
Trends in prevalence of HIV, syphilis, hepatitis C, hepatitis B, and sexual risk behavior among men who have sex with men. Results of 3 consecutive respondent-driven sampling surveys in Beijing, 2004 through 2006
J Acquir Immune Defic Syndr
Respondent-driven sampling in a study of drug users in New York City: notes from the field
J Urban Health
Methods to recruit hard-to-reach groups: comparing two chain referral sampling methods of recruiting injecting drug users across nine studies in Russia and Estonia
J Urban Health
From networks to populations: the development and application of respondent-driven sampling among IDUs and Latino gay men
AIDS Behav
HIV infection, sexual risk behavior, and substance use among Latino gay and bisexual men and transgender persons
Am J Public Health
Recruiting injection drug users: a three-site comparison of results and experiences with respondent-driven and targeted sampling procedures
J Urban Health
Exploring barriers to ‘respondent driven sampling’ in sex worker and drug injecting sex worker populations in eastern Europe
J Urban Health
Cited by (49)
An application of agent-based modeling to explore the impact of decreasing incarceration rates and increasing drug treatment access on sero-discordant partnerships among people who inject drugs
2021, International Journal of Drug PolicyCitation Excerpt :We parameterized the ABM using data from PWID who were recruited using RDS but did not account for this sampling procedure in analysis to maintain parsimony of the model. Prior research documents inconsistencies in the extent to which RDS sampling methods result in similar or different distribution of demographic characteristics, including age, racial/ethnic background, HIV status and residential location as compared to other recruitment methods (Abdul-Quader et al., 2006; Burt & Thiede, 2014; Burt, Hagan, Sabin, & Thiede, 2010; Collier, Garfein, Cuevas-Mota, & Teshale, 2017; Robinson et al., 2006; Rudolph, Young, & Lewis, 2015). Because the sample of PWID used for parameterizing the model exclusively lived in urban centers across the United States, this study's findings lacked generalizability to PWID living in rural and suburban settings.
Associations of place characteristics with HIV and HCV risk behaviors among racial/ethnic groups of people who inject drugs in the United States
2016, Annals of EpidemiologyCitation Excerpt :However, by adjusting for place and sociodemographic factors, intra-chain clustering may have been partially accounted for. Prior research suggests recruitment chains established from RDS are influenced by geography and demographic characteristics; thus, by controlling for geographic variance and individual-level demographic information, we may have partly accounted for potential bias established through RDS [86–88]. Log binomial and modified Poisson models did not converge to generate estimates of prevalence ratios in this sample in which both outcomes were prevalent.
Heroin-related overdose: The unexplored influences of markets, marketing and source-types in the United States
2015, Social Science and MedicineCitation Excerpt :Although our goal was to recruit a ‘typical street based’ user sample in both cities, our sampling was not random and bias may have been introduced. Respondent-Driven Sampling aims to yield a representative sample from small samples and while validated for some studies and populations may have limitations in reaching the most stigmatized and hidden populations (Rudolph et al., 2011; Burt et al., 2010). Sampling biases could also result from the targeted recruitment of the most socially excluded population of homeless or precariously housed heroin injectors as their methods of acquiring information about heroin strength may differ from those with more resources.
Challenges to recruiting population representative samples of female sex workers in China using Respondent Driven Sampling
2015, Social Science and MedicineCitation Excerpt :PLACE is a venue based sampling approach which focuses on the systematic identification of places where people meet new sexual partners (Weir et al., 2003). Unlike previous studies comparing RDS with alternative sampling approaches (Kendall et al., 2008; Robinson et al., 2006; Burt et al., 2010; Kral et al., 2010), the PLACE and RDS surveys were concurrently implemented, had the same inclusion definition and survey protocols. The PLACE-RDS comparison study found some differences in socioeconomic characteristics and prevalence of a positive syphilis test across sampling method, although it is not clear whether differences were due to violations of assumptions underlying RDS, distributional properties of the estimators, or a combination of both (Weir et al., 2012).
Current affiliations: New York University, New York City (H.H.), and the World Health Organization, Geneva, Switzerland (K.S.).