DESCRIPTION: This was a sub-study of the project “HIV/STD Risk Behaviors in Methamphetamine User Networks”. The purpose this project was to investigate epidemiological clustering of HIV as revealed by viral phylogenetic analysis to support or refute epidemiological linkage of pairs of HIV infected individuals, to describe broad patterns of clustering of infections by ethnic group and by structural features, and to describe the epidemiology of drug resistant HIV within this population
OUTCOMES: In our initial findings, the frequency of high level drug resistance was over 40% in individuals with viral loads > 500 copies/ml. This suggests that in this sample, treatment response is highly polarized. Analyses suggest relatively little clustering of HIV sequences, despite HIV+ individuals having an extremely high odds (>25%) of recruiting other HIV+ individuals, suggesting that there is little overlap between HIV transmission networks and recruitment by respondent-driven sampling. This also is consistent with HIV+ individuals reporting not having sex with their recruiter. This supplement helped to develop a statistical package, written in the R programming language, which imports RDS data, and generates figures depicting the recruitment networks, and generates sample, equilibrium and population estimates of quantities of interest, such as HIV prevalence.