RESUMO
BACKGROUND/AIMS: Anti-HCV positivity suggests past or present infection of HCV, or false positivity. The positive predictability of this test can differ according to the subjects. This study examines the positive predictability of the third generation anti-HCV ELISA and factors predicting HCV infection with special emphasis on the significance of the anti-HCV sample/cut-off (S/CO) ratio. METHODS: One hundred and ninety patients who were anti-HCV positive were enrolled, from November 1998 to January 2002 in Kyung Hee University Hospital. RT-PCR was performed to confirm HCV infection. RESULTS: One hundred and seven patients were RT-PCR positive (56.3% positive predictability). The positive predictability changed with the S/CO ratio: 17.9% in cases below 6, 58.3% between 6 and 50, 78.6% between 51 and 75, and 60% over 75. Those with the S/CO ratio more than 6 showed significantly higher predictability, but it did not increase further when the ratio got higher. Factors predicting HCV infection were the presence of liver cirrhosis (OR 5.5, p=0.000), hepatocellular carcinoma (OR 11.67, p=0.004), liver diseases (OR 2.99 p=0.001), and increase of AST (OR 2.49, p=0.002), ALT (OR 2.32, p=0.005), alpha-FP (OR 3.49, p=0.040), and the S/CO ratio of more than 6 (OR 7.82, p=0.000). However, liver cirrhosis was the sole factor in multivariate analysis (OR 8.32, p=0.02). CONCLUSIONS: The positive predictability of the third generation anti-HCV test was 56.3% with a significant difference between those with the S/CO ratio below 6 (18%) and above 6 (63%). In liver cirrhosis, positive predictability of anti-HCV test was relatively high as 85%.