Letter to the Editor: Regarding 'Individual and Neighborhood Correlates of Membership in Drug-Using Networks With a Higher Prevalence of HIV in New York City (2006-2009)'

摘要

Rudolph et al. [1] present data on the distribution of human immunodeficiency virus (HIV) infection in a respondent-driven sampling (RDS) recruitment network of drug users in New York City. Their article highlights the importance of integrating network and spatial perspectives into the study of HIV in marginalized populations. The effort is commendable as one of the first articles to explore network effects beyond the level of dyadic or ego-centric analysis using only standard RDS recruitment data. There are, however, several issues worth considering that challenge the authors' conclusions. We discuss analytical limitations and highlight key challenges that should be addressed when using RDS data. Resolving these issues will ultimately advance our understanding of the role of networks in HIV transmission.