Abstract
Respondent-driven sampling (RDS) is a relatively new form of sampling hidden populations. RDS has the ability to produce validpopulation estimates, provided that certain assumptions are met. This study examined whether an RDS survey of people who inject drugs (PWID) conducted in South East Sydney was able to meet the assumptions and requirements necessary to produce valid population estimates. From November 2009 to March 2010, 261 participants were recruited in 16 waves. While the method’s requirements and assumptions were largely met, an apparent lack of random recruitment and the low proportion of coupons returned were sources of concern. This study shows that RDS can be used successfully in Australia as a means of recruiting PWID. However, more work is needed to confirm the soundness of the assumptions on which the calculation of valid population estimates are based.