Prevalence of hepatitis C in tropical communities: the importance of confirmatory assays.

The prevalence of antibody to hepatitis C virus (HCV) was estimated in 3 tropical populations using 2 screening ELISAs to detect antibody to the c100-3 antigen and 2 supplementary assays designed to test the specificity of these tests. Two hundred and eighty-six of 385 (74.2%) sera from Kiribati, 17 of 138 (12.3%) sera from Vanuatu, and 39 of 173 (22.5%) sera from Zaire were reactive in the initial screening assay. The proportion of reactive sera which were also reactive in the second screening ELISA varied between populations (55.1% in Kiribati, 85.1% in Vanuatu, and 39.2% from Zaire). Reactive sera were selected at random for confirmatory testing. Only 3 of 49 (6.12%) of sera from Kiribati and 1 of 14 (4.76%) of sera from Vanuatu positive in the initial ELISA were reactive in the confirmatory assays. The proportion of confirmed positive sera from Zaire was higher 8 of 28 (28.5%). Based on the results of these supplementary assays the estimated prevalence of anti-HCV in these populations is 4.8% in Kiribati, less than 1% in Vanuatu, and 6.4% in Zaire. Reliance on a single screening ELISA to estimate the prevalence of anti-HCV in stored sera from tropical communities may lead to a gross over-estimate of the true prevalence in these populations.