Text this: Suboptimal reliability of liver biopsy evaluation has implications for randomized clinical trials.