Interobserver variability in transient elastography analysis of patients with chronic hepatitis C

Background & AimsTransient elastography based on liver stiffness measurement is a non-invasive method to assess hepatic fibrosis. However, interobserver variability has led to controversy over its use in fibrosis evaluation. To evaluate the interobserver variation in transient elastography in chronic hepatitis C. MethodsWe performed a cross-sectional study, analysing findings from two experienced operators who each assessed 195 patients by transient elastography on the same day. Liver stiffness measurement used to define fibrosis stages, based on METAVIR score, was: <7.1 as F0F1, 7.1-9.4 as F2, 9.5-12.4, as F3 and >12.4kPa as F4. We also assessed interobserver variation in identification of potential oesophageal varices screening based on transient elastography. ResultsThe interobserver intraclass correlation coefficient was 0.940 (95% CI 0.863-0.967) and measurements made by operators correlated [Spearman's =0.924; P<0.001]. However, the median liver stiffness measurement assessed by first operators was higher (11.5 vs 9.8kPa; P<0.001). The discordance between operators was 35% for at least one stage of fibrosis and 5% for two or more stages. Interobserver reliability values were =0.61 for fibrosis stages F2 and =0.80 for cirrhosis. Among the 74 patients determined to have cirrhosis by at least one operator, there was considerable discordance in identification of those with indication for oesophageal varices screening ( values from 0.13 to 0.61) according to several cut-offs. ConclusionAlthough a high correlation of liver stiffness measurement between operators, interobserver variability in transient elastography was not negligible. This method should not be used as the only screening tool for oesophageal varices in chronic hepatitis C.