Send to

Choose Destination
JMIR Public Health Surveill. 2019 Mar 14;5(1):e11737. doi: 10.2196/11737.

Estimating the Population Size of Female Sex Workers in Namibia Using a Respondent-Driven Sampling Adjustment to the Reverse Tracking Method: A Novel Approach.

Author information

Center for AIDS Prevention Studies, Division of Prevention Science, University of California, San Francisco, San Francisco, CA, United States.
Strategic Information/Monitoring and Evaluation, WHO India Country Office, Gurugaon, Haryana, India.
Directorate of Special Programmes, Response Monitoring & Evaluation Subdivision, Ministry of Health and Social Services, Windhoek, Namibia.
Division of Global HIV and Tuberculosis, US Centers for Disease Control and Prevention, Atlanta, GA, United States.
Institute for Global Health Sciences, University of California San Francisco, San Francisco, CA, United States.
Department of Epidemiology and Biostatistics, University of California, San Francisco, San Francisco, CA, United States.
Contributed equally



Key populations, including female sex workers (FSWs), are at a disproportionately high risk for HIV infection. Estimates of the size of these populations serve as denominator data to inform HIV prevention and treatment programming and are necessary for the equitable allocation of limited public health resources.


This study aimed to present the respondent-driven sampling (RDS) adjusted reverse tracking method (RTM; RadR), a novel population size estimation approach that combines venue mapping data with RDS data to estimate the population size, adjusted for double counting and nonattendance biases.


We used data from a 2014 RDS survey of FSWs in Windhoek and Katima Mulilo, Namibia, to demonstrate the RadR method. Information from venue mapping and enumeration from the survey formative assessment phase were combined with survey-based venue-inquiry questions to estimate population size, adjusting for double counting, and FSWs who do not attend venues. RadR estimates were compared with the official population size estimates, published by the Namibian Ministry of Health and Social Services (MoHSS), and with the unadjusted RTM.


Using the RadR method, we estimated 1552 (95% simulation interval, SI, 1101-2387) FSWs in Windhoek and 453 (95% SI: 336-656) FSWs in Katima Mulilo. These estimates were slightly more conservative than the MoHSS estimates-Windhoek: 3000 (1800-3400); Katima Mulilo: 800 (380-2000)-though not statistically different. We also found 75 additional venues in Windhoek and 59 additional venues in Katima Mulilo identified by RDS participants' responses that were not detected during the initial mapping exercise.


The RadR estimates were comparable with official estimates from the MoHSS. The RadR method is easily integrated into RDS studies, producing plausible population size estimates, and can also validate and update key population maps for outreach and venue-based sampling.


human immunodeficiency virus; population density; sex workers; social networking; vulnerable populations

Free full text

Supplemental Content

Full text links

Icon for JMIR Publications
Loading ...
Support Center