Prediction of compensated liver cirrhosis by ultrasonography and routine blood tests in patients with chronic viral hepatitis

BACKGROUND/AIMS Liver biopsy is a standard method for diagnosis of liver cirrhosis in patients with chronic hepatitis. Because liver biopsy is an invasive method, non-invasive methods have been used for diagnosis of compensated liver cirrhosis in patients with chronic hepatitis. The current study was designed to evaluate the usefulness of ultrasonography and routine blood tests for diagnosis of compensated liver cirrhosis in patients with chronic viral hepatitis. METHODS Two hundred three patients with chronic viral hepatitis who underwent liver biopsy were included in this study and ultrasonography and routine blood tests were analyzed retrospectively. Ultrasonographic findings, including surface nodularity, parenchyma echogenecity, and spleen size, were evaluated. The diagnostic accuracy of ultrasonography and routine blood tests were examined. RESULTS Discriminant analysis with forward stepwise selection of variables showed that liver surface nodularity, platelet count, and albumin level were independently associated with compensated liver cirrhosis (p<0.05). Cross-tabulation revealed that the following 4 variables had >95% specificity: platelet count <100,000 /uL; albumin level <3.5 g/dL; INR >1.3; and surface nodularity. If at least one of the four variables exists in a patient with chronic viral hepatitis, we can predict liver cirrhosis with 90% specificity and 61% sensitivity. CONCLUSIONS These results suggest that four variables (platelet count <100,000 /uL, albumin level <3.5 g/dL, INR >1.3, and surface nodularity) can be used for identification of liver cirrhosis in patients with chronic viral hepatitis with high specificity.