Skepticism about rds inference validity: Preliminary empirical findings from a recruitment dynamic and network study

Background: Respondent Driven Sampling (RDS) has gained its popularity internationally in HIV research and epidemiological surveillance for its cost effectiveness in reaching hidden populations and claimed ability to make unbiased population estimates. However, RDS inference models were based on strong but unsupported assumptions regarding peer recruitment processes and the structure of underlying social networks. The inference validity can be threatened when one or more of the underlying assumptions fail. Methods: A mixed method intensive social network study is being conducted in Hartford, CT to recruit a typical RDS sample of 500 injection drug users (IDUs). Comprehensive social network surveys before and after actual peer recruitment, field outreach, as well as 60 in-depth interviews will be analyzed to assess IDUs' peer recruitment intentions, actual experiences, and real world contextual factors associate with their multi-layered social network composition and structures related to RDS implementation. Results: Preliminary ego-network analysis does not support random peer recruitment assumptions. IDUs were more likely to pass recruitment coupons to strong ties, those who need money, reliable for appointments, depend more on drugs, and those within convenient social and geographical distances. The role and strength of contextual factors shaping peer selection decision making, recruitment experience and enrollment success were different, although previous inference models failed to differentiate these dynamic behaviors and associated network compositions. Implications of these patterns to the validity of RDS inferences will be discussed. Conclusions: Empirical network data specifically related to RDS implementation and recruitment dynamics are necessary to improve RDS estimators in the future.