Article Text

Download PDFPDF

Epidemiology poster session 4: Methodological aspects
P1-S4.08 Population size estimates for men who have sex with men in Guatemala City using time location sampling and respondent driven sampling
  1. G Paz-Bailey1,
  2. B Alvarez2,
  3. W Miller2,
  4. B Sabrina3,
  5. C Barrington4,
  6. A Kim5,
  7. S Morales2,
  8. S Chen5
  1. 1Tephinet Inc, Guatemala, Guatemala
  2. 2Del Valle University of Guatemala, Guatemala, Guatemala
  3. 3Tephinet Inc, Guatemala
  4. 4University of North Carolina, USA
  5. 5Centers for Disease Control and Prevention, USA


Background Men who have sex with men (MSM) are highly vulnerable to HIV infection, but this population can be particularly difficult to reach in Central America due to stigma and violence. Knowing the size of populations at high risk for HIV and sexually transmitted infections (STI) is critical for informing prevention, care, and treatment programs. Simple approaches are needed to provide baseline estimates of the population size of MSM in Guatemala to advocate for appropriate resource allocation and programming. We compared population size estimates of MSM and transgender persons using capture-recapture linked to two differentsurveys that were developed concurrently using different sampling methodologies.

Methods Capture recapture methods for estimating population size were integrated into two probability-based surveys using respondent driven sampling (RDS) and time-location sampling (TLS); conducted simultaneously among MSM in Guatemala City from October to December 2010. Key chains were used as uUnique objects in the form ofand study key chains were distributed approximately 1-month prior to the surveys in venues known to be frequented by MSM. Duplicate participation was avoided by using the same team to distribute the key chains and administering a set of questions before handing the object. Participants in the RDS and TLS surveys were asked, as part of the study interview, whether they had received the key chain. We assigned the number of key chains distributed in venues as the first capture and the proportion of participants that reported receiving the key chain in the RDS or TLS survey as the second capture. By applying standardised formulas for estimating population size using capture-recapture methods, we estimated the number of MSM in Guatemala City and 95% CIs around this estimate, adjusted for RDS and TLS sampling design.

Results A total of 2128 key chains were distributed in the first capture. Of the 501 MSM interviewed in the RDS survey, 200 (RDS adjusted, 23.4%) had received the key chain. Thus the RDS adjusted estimated population size of MSM in Guatemala City was 9,190 (95% CI 7765 to 10 616). Of the 504 MSM interviewed in the TLS survey, 193 (TLS adjusted, 32.1%) had received the key chain. The TLS adjusted estimated number of MSM was 6620 (95% CI 5813 to 7427).

Conclusions Estimates of MSM population size using the TLS survey resulted in a significantly lower estimate than that obtained through RDS survey, probably due to the fact that TLS targets MSM who attend venues. The estimate obtained through RDS may better reflect the size of the larger MSM population in Guatemala City, since this itRDS reaches MSM attending and not attending publicmeeting venues. Integrating capture-recapture methods in probability-based surveys among MSM is a simple and fast approach for providing the population sizes needed to inform and evaluate programs for MSM in Guatemala.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.