Transfer of sampling methods for studies on most-at-risk populations (MARPs) in Brazil

Departamento de DST, AIDS e Hepatites Virais, Ministério da Saúde, Brasília, Brasil.
Cadernos de saúde pública / Ministério da Saúde, Fundação Oswaldo Cruz, Escola Nacional de Saúde Pública (Impact Factor: 0.98). 01/2011; 27 Suppl 1(Suppl 1):S36-44. DOI: 10.1590/S0102-311X2011001300005
Source: PubMed


The objective of this paper was to describe the process of transferring two methods for sampling most-at-risk populations: respondent-driven sampling (RDS) and time-space sampling (TSS). The article describes steps in the process, the methods used in the 10 pilot studies, and lessons learned. The process was conducted in six steps, from a state-of-the-art seminar to a workshop on writing articles with the results of the pilot studies. The principal investigators reported difficulties in the fieldwork and data analysis, independently of the pilot sampling method. One of the most important results of the transfer process is that Brazil now has more than 100 researchers able to sample MARPs using RDS or TSS. The process also enabled the construction of baselines for MARPS, thus providing a broader understanding of the dynamics of HIV infection in the country and the use of evidence to plan the national response to the epidemic in these groups.

Download full-text


Available from: Carl Kendall,
  • Source
    • "Respondent-driven sampling (RDS) is a recently introduced link-tracking network sampling technique for collecting such information (Heckathorn, 1997). Because of the pressing need for information about the most at-risk groups and the weaknesses of alternatives approaches, RDS has already been used in more than 120 HIV-related studies in 20 countries (Malekinejad et al., 2008) and has been adopted by leading public health organizations, such as the US Centers for Disease Control and Prevention (CDC) (Barbosa Júnior et al., 2011; Lansky et al., 2007; Montealegre et al., 2012). "
    [Show abstract] [Hide abstract]
    ABSTRACT: Respondent-driven sampling (RDS) is a widely used method for sampling from hard-to-reach human populations, especially groups most at-risk for HIV/AIDS. Data are collected through a peer-referral process in which current sample members harness existing social networks to recruit additional sample members. RDS has proven to be a practical method of data collection in many difficult settings and has been adopted by leading public health organizations around the world. Unfortunately, inference from RDS data requires many strong assumptions because the sampling design is not fully known and is partially beyond the control of the researcher. In this paper, we introduce diagnostic tools for most of the assumptions underlying RDS inference. We also apply these diagnostics in a case study of 12 populations at increased risk for HIV/AIDS. We developed these diagnostics to enable RDS researchers to better understand their data and to encourage future statistical research on RDS.
    Journal of the Royal Statistical Society Series A (Statistics in Society) 09/2012; 178(1). DOI:10.1111/rssa.12059 · 1.64 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Female sex workers are known in Brazil and elsewhere in the world as one of the most-at-risk populations for risk of HIV infection, due to their social vulnerability and factors related to their work. However, the use of conventional sampling strategies in studies on most-at-risk subgroups for HIV is generally problematic, since such subgroups are small in size and are associated with stigmatized behaviors and/or illegal activities. In 1997, a probabilistic sampling technique was proposed for hard-to-reach populations, called Respondent-Driven Sampling (RDS). The method is considered a variant of chain sampling and allows the statistical estimation of target variables. This article describes some assumptions of RDS and all the implementation stages in a study of 2,523 female sex workers in 10 Brazilian cities. RDS proved appropriate for recruiting sex workers, allowing the selection of a probabilistic sample and the collection of previously missing information on this group in Brazil.
    Cadernos de saúde pública / Ministério da Saúde, Fundação Oswaldo Cruz, Escola Nacional de Saúde Pública 01/2011; 27 Suppl 1(Suppl 1):S45-55. DOI:10.1590/S0102-311X2011001300006 · 0.98 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Respondent-driven sampling (RDS) is a chain-referral method that is being widely used to recruit most at-risk populations. Because the method is respondent driven, observations are dependent. However, few publications have focused on methodological challenges in the analysis of data collected by RDS. In this article, we propose a method for estimating the variance of the HIV prevalence rate, based on the Markov transition probabilities and within recruitment cluster variation. The method was applied to a female commercial sex workers study carried out in 10 Brazilian cities in 2008. Both the inverse of network size and the size of the city were considered in the estimation of overall sampling weights. The study included a behavior questionnaire and rapid tests for HIV and syphilis. About 2523 interviews were conducted successfully, excluding the seeds. Results show a positive homophily between recruits for those HIV+; HIV- recruiters selected HIV+ recruits 4% of the time; HIV+ recruiters selected other HIV+ recruits 19.6% of the time, about 5 times higher. The prevalence rate was estimated at 4.8% (95% confidence interval: 3.4 to 6.1), and a design effect of 2.63. Using statistical methods for complex sample designs, it was possible to estimate HIV prevalence, standard error, and the design effect analytically. Additionally, the proposed analysis lends itself to logistic regression, permitting multivariate models. The stratification in cities has proved suitable for reducing the effect of design and can be adopted in other RDS studies, provided the weights of the strata are known.
    JAIDS Journal of Acquired Immune Deficiency Syndromes 08/2011; 57 Suppl 3(Suppl 3):S129-35. DOI:10.1097/QAI.0b013e31821e9a36 · 4.56 Impact Factor
Show more