Can reference images improve interobserver agreement in reporting liver fibrosis?

Staging of fibrosis in medical liver biopsies has inherent interobserver variability. There are a number of disease-specific scoring systems available. While recognising the importance of these scoring systems, there is scope to consider how concordance amongst histopathologists could be improved using a generic fibrosis staging system.Using virtual slides, we approached both specialist liver histopathologists and general histopathologists from the UK to assess the degree of fibrosis against a proposed four-tiered reporting system. Example reference images were then produced and distributed to the same responders who were asked to rate a second set of slides to assess if the use of reference images improved concordance between pathologists.The use of reference images eliminated spread across three categories (from 15% to 0%). Overall, agreement was already good; our study showed an improved agreement amongst all participants for percentage agreement (67.79% to 70.08%) and interobserver agreement improved (Fleiss' Kappa 0.55 to 0.59).