In Reply: Drs Hallett and Garnett and Mr Todd and colleagues raise legitimate concerns regarding the accuracy of the BED assay in estimating HIV incidence. The assay was developed to include all subtypes and recombinants, including those in Africa, but has previously overestimated HIV incidence and requires an adjustment factor to address misclassification.1 For this reason, we focused our analysis on risk factors for recent infection. A large proportion of BED-misclassified infections occur in the year prior to testing,2 which is recent from a risk factor standpoint.
Mermin J, Musinguzi J, Hladik W. Estimating Incidence of HIV Infection in Uganda—Reply. JAMA. 2009;301(2):159-161. doi:10.1001/jama.2008.956