Evidence from National Population-Based Surveys on Bias in Antenatal Clinic-Based Estimates of HIV Prevalence

Simona Bignami, Harvard University
Joshua A. Salomon, Harvard University
Christopher J.L. Murray, Harvard University

National HIV estimates in most developing countries with generalized epidemics (defined as a prevalence of at least 1% in the general population) are based on data generated by surveillance systems that focus on pregnant women attending a selected number of antenatal clinics (ANC). However, an increasing number of community-based studies have revealed the limitations inherent in antenatal clinic-based surveillance data for correctly estimating adult female and male HIV prevalence in the general population. In this paper, we provide a comprehensive assessment of the bias in national estimates of adult HIV prevalence from ANC-based surveillance, by using data from recent nationally representative, population-based surveys in Mali, Zambia, Dominican Republic, Burundi, Niger, South Africa, Kenya, Burkina Faso, and Ghana. We assess and adjust for non-response bias in the survey data, and then compare HIV prevalence in urban and rural areas between the survey and the appropriate ANC data.