Modeling Power to Detect Changes in HIV Incidence Through Routine ANC Attendance Data

• Activity Number: 379 - Understanding HIV/AIDS Epidemics with Newly Available Data Sources
• Type: Invited
• Date/Time: Tuesday, August 1, 2017 : 2:00 PM to 3:50 PM
• Sponsor: Section on Statistics in Epidemiology
• Author(s): Britta J. Jewell* and Jennifer Smith and Timothy B. Hallett and Ann Duerr
• Companies: UC Berkeley and Imperial College London and Imperial College London and Fred Hutchinson Cancer Research Center
• Keywords: HIV ; mathematical models ; incidence estimation

Abstract

Measuring HIV incidence over time is imperative for monitoring changes in epidemics. We developed an individual-based mathematical model of a population of adolescent girls and young women (AGYW) eligible to acquire HIV-1, become pregnant, and attend antenatal clinics (ANC) during pregnancies. The model was parameterized for four separate settings in sub-Saharan Africa (South Africa, Tanzania, Uganda, and Zimbabwe). Using different strategies of incidence estimation (changes in prevalence as a proxy for incidence, a synthetic cohort strategy, and use of a LAg-Avidity incidence assay), we estimated the statistical power of each strategy to detect either a 25% reduction in HIV incidence over one year or a 40% reduction incidence over two years. Results will be discussed, including statistical power for each incidence measurement strategy and the use of individual-based mathematical models for performing power calculations.