Time location sampling and respondent driven sampling: techniques implementation for monitoring concentrated HIV/AIDS epidemic in Mexico

Cecilia Gayet, Facultad Latinoamericana de Ciencias Sociales (FLACSO)
Araceli Fernández-Cerdeño, University of California, San Diego

Background: There are innovative sampling methods for hard to reach populations. A Mexican team surveyed male and female sex workers (MSW and FSW), men who have sex with men (MSM), and long-distance truck-drivers (truckers) in four cities using Time-Location Sampling (TLS) and Respondent-Driven Sampling (RDS). Objectives: To share lessons learned and discuss feasibility, advantages and challenges of TLS and RDS sampling for the study of hard-to-reach populations in low-resource settings. Methods: TLS samples included FSW (n=603), MSM (n=1,111) and truckers (n=313). RDS was used for MSW (N=102). We present an analysis of quantifiable and unquantifiable biases based on field conditions, quality assurance procedures, methods’ limitations and rejection rates. Results: Expected sample were achieved for 3 out of 4 target groups. Selection bias was smaller than with convenience sampling methods. Discussion: We discuss randomness of achieved samples, quantifiable and non-quantifiable biases and suggest ways to minimize arbitrary selection in the field.