International Journal of Epidemiology

Human immunodeficiency virus (HIV) is one of the most crucial problems in Africa. Prevalence of HIV is very high in many African countries. For this reason, it is important to have a tool that allows us to give reliable and definitive HIV serology test results. Western blotting is one of the reference confirmatory tests for the diagnosis of HIV infection or after inconclusive EIA results. However, the use of WB is limited by difficulty in their interpretation. Two different diagnostic criteria have already been proposed by the World Health Organization (WHO)1 and the Centers for Disease Control and Prevention (CDC).2 This difficulty, and cost, has led to a reduction in recent years in the use of WB by groups such as WHO which recommend using a combination of EIA which are easier to interpret. Nevertheless, WB tests are still commonly used to confirm results from these EIA. In Africa, the presence of non-specific bands on WB lead to many tests being labelled indeterminate, undermining the use of WB as a diagnostic tool.3–5 There is no clear biological explanation as to why the host response on WB bands differs in Africans compared to others. Several hypotheses have been advanced including cross-reactivity with malaria, other retro-viruses, Herpes simplex virus infection, HIV-2, group O viruses, and anti-HLA antibodies.6–9 In this paper, we have investigated the possibility of develop-ing simple and reliable diagnostic criteria for the interpretation of confirmatory WB in epidemiological studies in Africa. Specific-ally, our objective was to determine which set of WB bands are the most predictive of HIV status in a large Ugandan population and to establish appropriate diagnostic criteria, which minimize