Abstract:
|
We describe the development of statistical models to incorporate HIV prevalence data collected through routine testing (RT) for HIV of pregnant women attending antenatal clinics (ANCs). This new RT data source, in addition to historical estimates of HIV prevalence from ANC unlinked anonymous testing as part of periodic sentinel surveillance (SS) and nationally-representative population-based surveys (NPS) are used as the main inputs for monitoring the trajectory of the HIV epidemic. We propose new statistical models for using RT data at the health facility level ('site-level') and aggregated regionally ('census-level'), depending on data availability. We use a synthetic RT data set to measure their potential for impacting overall prevalence trends.
|