Background: Transient elastography by fibroscan is widely used as non-invasive marker for liver fibrosis. Fibroscan has limited accuracy in patients with elevated serum bilirubin, ALT, ascites, BMI >30 kg/m2 and requires separate equipment for evaluation. ARFI is a new way of assessing liver fibrosis which is incorporated in ultrasound machine. In a prospective study we evaluated utility of ARFI in 171 consecutive patients undergoing liver biopsy and assessed misclassification rate of fibrosis in patients with elevated serum bilirubin, ALT, BMI >25 kg/m2 and ascites. 

Methods: 171 consecutive patients (mean age 46.04±13.34 years, males 60%) undergoing liver biopsy were evaluated for ARFI by Acuson S2000, Siemens ultrasonography machine. Indications for liver biopsy were autoimmune liver disease (38), alcoholic liver disease (13), non-alcoholic fatty liver disease (24), cryptogenic liver disease (10), noncirrhotic portal hypertension (29), viral hepatitis (17) and miscellaneous (40). Liver biopsy was graded using Metavir classification. Aspartate aminotransferase to platelet ratio index (APRI) was calculated in each patient. ARFI and APRI values were correlated with fibrosis stage on liver biopsy. ARFI and fibrosis were compared with respect to serum bilirubin, ALT, BMI >25 kg/m2 and ascites. 

Results: Mean ARFI score with interquartile range were 1.70±0.46 (1.33-1.99) in stage 0, 1.92±0.56 (1.45-2.16) in stage 1-2 and 2.49±0.60 (2.07-2.98) in stage 3 and 4 fibrosis. Mean ARFI and APRI values were significantly higher in stage 3 and 4 fibrosis as compared to fibrosis stage 0 and 1-2 (p 0.001). ARFI values were misclassified in patients with bilirubin > 10 mg/dl, ALT > 300 IU/L and ascites. 

Conclusion: ARFI was reliable predictor of advanced fibrosis but could not differentiate stage 0 from early fibrosis. ARFI values were not reliable in bilirubin >10 mg/dl, ALT > 300 IU/L and ascites.