Our analysis showed that culture or Xpert were cost-effective at reducing early mortality during the first 6 months of ART compared with the current practice. Thus, our findings provide support for ongoing efforts to expand TB diagnostic capacity.

We present a comparison of respondent-driven sampling (RDS) and time-location sampling (TLS) for behavioral surveillance studies among men who have sex with men (MSM). In 2010, we conducted two simultaneous studies using TLS (N = 609) and RDS (N = 507) in Guatemala city. Differences in characteristics of the population reached based on weighted estimates as well as the time and cost of recruitment are presented. RDS MSM were marginally more likely to self-report as heterosexual, less likely to disclose sexual orientation to family members and more likely to report sex with women than TLS MSM. Although RDS MSM were less likely than TLS MSM to report ≥2 non-commercial male partners, they were more likely to report selling sex in the past 12 months. The cost per participant was $89 and $121 for RDS and TLS, respectively. Our results suggest that RDS reached a more hidden sub-population of non-gay-identifying MSM than TLS and had a lower implementation cost.