OBJECTIVESTo evaluate the correlation between signal/cutoff (S/CO) ratios of anti-HCV EIA and their true positivity for determining the predictive value of S/CO ratios.

METHODSOne hundred and fifty-nine samples of blood from donors positive for anti-HCV at the initial screening were collected from Beijing, Guangzhou, Hangzhou, Kunming and Urumchi. All the samples were retested by Ortho and 6 Chinese domestic anti-HCV EIA kits in duplicate, and detected for HCV RNA (NAT) using Chiron Procleix HIV/HCV system (transcription mediated amplification, TMA). The HCV RNA negative samples were further tested for anti-HCV by Chiron RIBA 3.0. Either NAT or RIBA positive samples were interpreted as the true positive.

RESULTSAll 7 anti-HCV EIA kits had a significant correlation between S/CO ratios and true positivity. The S/CO ratio of Ortho > or = 3.8 predicted the true positivity in 96.1% of the samples tested. The S/CO ratios of BGI-GBI, GWK, SABC, KHB, InTec, and Wantai were > or = 7.0, > or = 10.0, > or = 6.0, > or = 10.0, > or = 8.6, > or = 14.0 and predicted 96.1%, 96.1%, 97.3%, 96.0%, 96.1%, 96.0% of the true positivity, respectively.

CONCLUSIONSThe S/CO ratios of anti-HCV EIA kits are associated with the true positivity. S/CO ratios of Ortho, BGI-GBI, GWK, SABC, KHB, InTec and Wantai predicting > or = 95% true positivity are > or = 3.8, > or = 7.0, > or = 10.0, > or = 6.0, > or = 1 0.0, > or = 8.6 and > or = 14.0, respectively.