Histopathologist Features Predictive of Diagnostic Concordance at Expert Level Amongst a Large International Sample of Pathologists Diagnosing Barrett′s Dysplasia Using Digital Pathology

Objective: Guidelines recommend expert pathology review of Barrett′s oesophagus (BO) biopsies that reveal dysplasia, but there are no evidence-based standards to corroborate expert reviewer status. We investigated BO concordance rates and pathologist features predictive of diagnostic discordance amongst a large international cohort of gastrointestinal pathologists to develop a quantitative model of BO expert review. Design: Pathologists (n=55) from over 20 countries assessed 55 digitised BO biopsies from across the diagnostic spectrum, before and after viewing matched p53 immunohistochemistry. Extensive demographic and clinical experience data were obtained via online questionnaire. We calculated discordance rates and applied multivariate regression analyses to identify predictors of concordance. Results: We recorded over 6,000 individual case diagnoses. Of 2,805 H&E diagnoses, we found excellent concordance (>70%) for non-dysplastic Barrett′s oesophagus (NDBO) and high-grade dysplasia (HGD), and intermediate concordance for low-grade dysplasia (LGD, 42%) and indefinite for dysplasia (IND, 23%). Major diagnostic errors (i.e. NDBO overinterpreted as LGD/HGD or vice versa) were found in 248 diagnoses (8.8%), which reduced to 8.3% after viewing p53 labelled slides. At least 5 years of professional experience was protective against major diagnostic error for H&E slide review (OR 0.48, 95%CI 0.31-0.74). Working in a district general hospital was associated with increased odds of major diagnostic error (OR 1.76, 95%CI 1.15-2.69), however this was neutralised when pathologists viewed p53 labelled slides, suggesting a beneficial impact of p53 immunohistochemistry for this group. Conclusion: We have developed an evidence-based quantitative model of BO histopathology diagnosis at expert consensus level that will inform guideline development.