Bayesian Approaches for Estimating Prevalence Based on Pool Screening When Observing Zero Positive Pools

• Activity Number: 508
• Type: Contributed
• Date/Time: Wednesday, August 3, 2011 : 10:30 AM to 12:20 PM
• Sponsor: Section on Bayesian Statistical Science
• Author(s): Thomas Birkner*+ and Immaculada Aban and Charles Katholi
• Companies: University of Alabama at Birmingham and University of Alabama at Birmingham and University of Alabama at Birmingham
• Keywords: Pool screening ; sequential Bayes ; objective prior

Abstract

Pool screening is a method that combines individual units into pools. Each pool will either test positive (at least one of the units is positive) or negative (all units are negative). Pool screening is commonly applied to the study of tropical diseases where pools consist of vectors that can carry the disease (e.g. black fly). The goal is to estimate the proportion of infected vectors. When control programs are effective and the surveillance phase is in place, entomologists may encounter data with no positive pools. The traditional maximum likelihood estimate of the prevalence is zero - an unrealistic value. A sequential Bayesian approach can incorporate results from previous years and will provide a more sensible prevalence estimate even if all of the pools collected at this time are negative for the trait of interest. This paper proposes and evaluates the performance of a Bayesian approach to this problem. Through simulation, we investigate the amount of data (pool size, number of years) required such that the type of objective prior chosen (e.g. Bayes/Laplace, Jeffreys) does not make a significant difference with respect to the estimate of prevalence.