Strategies for reliable diagnosis of hepatitis C infection: the need for a serological confirmatory assay.

The aim of the study was to examine whether the diagnosis of Hepatitis C (HCV) infection can be obtained reliably without using an immunoblot-based confirmation assay. 1,708 EIA-reactive serum samples were examined retrospectively for (i) optical density value in the screening assay, (ii) reactivity in an immunoblot assay, and (iii) result by RT PCR. In 1,394 (81.0%) samples positive results were obtained by both the HCV EIA and the confirmation assay. OD-values > or = 2.2 were observed in 1026 of these samples, but covered the range from 0.4 to 2.1 in the other 368 samples. The combination of HCV EIA reactivity and indeterminate immunoblot assay was observed in 134 (7.8%) serum samples. HCV RNA was detected in 58 cases by PCR. The OD-values of these 58 samples ranged from 0.4 to >2.2. Especially reactivity against the core recombinant protein was indicative of PCR positivity. The reactivity by the HCV EIA could not be confirmed by immunoblot assay or PCR in 180 (10.5%) sera. These false reactive sera showed OD values by EIA from 0.3 to 2.1. It is concluded that no threshold values can be defined which would allow differentiation between positive, indeterminate, and false reactive result by HCV EIA without producing an unacceptably high number of false negative diagnoses. Not using immunoblot-based confirmation would result in many additional PCR examinations. Therefore, confirmation of reactive HCV EIA results by a serological confirmatory assay must remain an essential part of the diagnostic procedure.