Short communication: Simplified estimation of the long-term specificity of the BED assay to improve estimates of HIV incidence

The BED assay overestimates HIV incidence because a proportion, epsilon (ω), of chronically infected people test "false-recent." In the ZVITAMBO trial, 2796 postpartum women tested HIV positive at baseline and 5.1% tested "false-recent" by BED 12 months later, providing a population-based estimate of ω. At baseline, 7.2% (95% CI: 5.3-9.4) of positive women >30 years and 5.1% (95% CI: 3.5-7.2) of positive women with CD4 <200 cells/μl tested "recent" by BED. Thus, the proportions of BED-positive women in either of these subgroups (data that would be available in surveys measuring HIV and BED status) provided a reasonably tight upper bound for ω, which can be used to provide a lower bound for HIV incidence.