Sampling hard-to-reach populations with respondent driven sampling

摘要

Cost effective and targeted prevention, intervention and treatment programs for hard-to-reach populations at risk for HIV and other infections rely on the collection of quality data through biological and behavioral surveillance surveys (BBSS). Over the past decade, there has been a global expansion of BBSS to measure the prevalence of HIV and other infections, and related risk behaviors among injecting drug users, males who have sex with males, and female sex workers. However, a major challenge to sampling these hard-to-reach populations is that they are usually stigmatised and/or practice illegal behaviors which, in turn, make them difficult to access and unwilling to participate in research efforts. Over the past decade, respondent driven sampling (RDS) has become recognised as a viable option for rigorous sampling of hard-to-reach populations. This paper introduces RDS methods and describes some of the advantages and challenges to implementing and analysing surveys that use RDS.