Effects of interventions on intra- and interobserver agreement on interpretation of nonalcoholic fatty liver disease histology

Accurate and reproducible interpretation of nonalcoholic fatty liver disease (NAFLD) histology has significant clinical and research-related implications. We evaluated the impact of 2 interventions ([1] review of illustrative histologic images of NAFLD with the study pathologists; [2] use of a scoring sheet with written diagnostic criteria for different NAFLD phenotypes) on intra- and interobserver agreement on interpretation of NAFLD histology. Before and after the interventions, 2 pathologists twice read 65 liver biopsies done for evaluation of suspected NAFLD. The intra- and interobserver agreement was highest on assessment of steatosis and fibrosis. The interventions significantly improved the intraobserver agreement only on assessment of hepatocellular ballooning. The interobserver agreement was only fair on assessment of lobular inflammation, ballooning, and diagnostic classification and did not improve after the interventions. Methods to improve interobserver agreement on assessment of lobular inflammation and ballooning are needed and would likely increase pathologists' agreement on NAFLD diagnostic classification.