Relying on injection drug users to access and adhere to HIV therapeutics: Bittersweet lessons using respondent-driven sampling and a peer-driven intervention

Results are presented of enrolling HIV+ active-injection drug users (IDUs) into a peer-driven intervention (PDI) to improve their adherence to medical care. Using respondent-driven sampling (RDS), which evolved out of the PDI model, the authors recruited and tested 1,097 IDUs, of whom 145 were confirmed to be HIV+. Despite promises of confidentiality, increased social supports, and direct cash rewards for participating, only 78 (54%) of the HIV+ IDUs enrolled in the subsequent adherence study. Correlates of participation as well as interviews with respondents who declined participating are presented. The seemingly negative findings have important implications for future adherence studies. RDS provided a powerful mechanism for recruiting HIV+ IDUs. But selection biases were revealed to favor sicker yet more socially connected respondents, which resulted in more conservative outcomes with regard to social supports and positive health-seeking behaviors. The analysis may help streamline future efforts using the PDI, which harnesses IDUs' abilities to help one another, by underscoring some of the mechanism's limitations.