Modeling the Marked Presence-Only Data: A Case Study of Estimating the Female Sex Worker Size in Malawi

• Activity Number: 1 - Invited E-Poster Session
• Type: Invited
• Date/Time: Sunday, August 2, 2020 : 12:30 PM to 3:30 PM
• Sponsor: Government Statistics Section
• Author(s): Ian Laga* and Le Bao and Xiaoyue Maggie Niu
• Companies: Penn State University and Penn State University and Penn State University
• Keywords: HIV; Bayesian modeling; Small area estimation; Prevalence mapping

Abstract

Continued fine-scale mapping of HIV/AIDS populations is needed to meet goals set by global organizations like UNAIDS and WHO. Since key populations like female sex workers (FSW), men who have sex with men (MSM), and people who inject drugs often have higher prevalence of HIV, it is of interest to map these specific key populations. However, key populations are typically difficult (or impossible) to map directly due to social stigma and legal issues, among other reasons. Instead, targeted surveys coupled with auxiliary data can be used to estimate population size with reasonable confidence intervals. One such study collected FSW data by interviewing a subset of venues believed to house FSW and obtained accurate estimates of district-level venue counts. These venues represent a spatial marked presence-only data set. No existing method can estimate the FSW abundance at a fine resolution. Thus, a calibrated-Bayesian model was developed to estimate the prevalence at a fine-cell-level. The number of FSW at approximately 1.5 by 1.5-km resolution and the corresponding credible intervals are estimated, providing a method for treating HIV in these key populations.