Bootstrap Variance and Bias Correction in the Estimation of HIV Incidence from Surveillance Data with Testing for Recent Infection

• Activity Number: 34
• Type: Contributed
• Date/Time: Sunday, August 2, 2009 : 2:00 PM to 3:50 PM
• Sponsor: Section on Statistics in Epidemiology
• Author(s): Nicole B. Carnegie*+
• Companies: University of Washington
• Keywords: HIV incidence ; model-based bootstrap ; testing for recent infection

Abstract

Karon et al. (2008) developed a model to estimate the incidence of HIV infection from surveillance data with biologic testing for recent infection. This method has been implemented by public health departments across the United States and is behind the new national incidence estimates, which are about 40% higher than previous estimates. We show that the delta method approximation given for the variance of the estimator is incomplete, leading to an inflated variance estimate. This contributes to the generation of overly conservative confidence intervals, potentially obscuring important differences between populations. We demonstrate via simulation that a model-based bootstrap using the specified model for data generation vastly improves variance estimation and also allows for bias correction of the point estimate, which could account for a portion of the change in the incidence estimates.