OBJECTIVE: To develop and validate a non-invasive index to predict the presence of cirrhosis in patients with chronic hepatitis C on the basis of clinical, laboratory, and ultrasound findings. MATERIALS AND METHODS: Data from the complete history and physical examination, serologic studies, liver ultrasound, and liver biopsy of patients with chronic hepatitis C were analyzed using multivariate regression to develop a cirrhosis predictive index. This index was then applied prospectively to another group of patients with chronic hepatitis C to determine its accuracy. RESULTS: Three hundred and thirty-two patients were included (mean age, 48.5+/-18.7 years; male-female ratio, 1.27). Sixty-seven patients (20%) had cirrhosis at histology. Logistic regression identified seven variables that predicted cirrhosis: age>or=60 years, platelet count<or=100 (x10/L), AST/ALT>or=1, prothrombin time (Ratio)>or=1.1, caudate hypertrophy, right lobe atrophy and splenomegaly. Patients scoring>or=22 in total had a statistically significant probability of cirrhosis (sensitivity, 80%; specificity, 96%; and diagnostic accuracy, 94%). CONCLUSION: Cirrhosis can be predicted in patients with chronic hepatitis C by the evaluation of seven clinical, laboratory, and sonographic variables. The index will be useful for the management and follow-up of hepatitis C patients drastically reducing the indications for biopsy in this context.
OBJECTIVE: To develop and validate a non-invasive index to predict the presence of cirrhosis in patients with chronic hepatitis C on the basis of clinical, laboratory, and ultrasound findings. MATERIALS AND METHODS: Data from the complete history and physical examination, serologic studies, liver ultrasound, and liver biopsy of patients with chronic hepatitis C were analyzed using multivariate regression to develop a cirrhosis predictive index. This index was then applied prospectively to another group of patients with chronic hepatitis C to determine its accuracy. RESULTS: Three hundred and thirty-two patients were included (mean age, 48.5+/-18.7 years; male-female ratio, 1.27). Sixty-seven patients (20%) had cirrhosis at histology. Logistic regression identified seven variables that predicted cirrhosis: age>or=60 years, platelet count<or=100 (x10/L), AST/ALT>or=1, prothrombin time (Ratio)>or=1.1, caudate hypertrophy, right lobe atrophy and splenomegaly. Patients scoring>or=22 in total had a statistically significant probability of cirrhosis (sensitivity, 80%; specificity, 96%; and diagnostic accuracy, 94%). CONCLUSION:Cirrhosis can be predicted in patients with chronic hepatitis C by the evaluation of seven clinical, laboratory, and sonographic variables. The index will be useful for the management and follow-up of hepatitis C patients drastically reducing the indications for biopsy in this context.