Objectives: The purpose of this study was to test the hypothesis that clinical information submitted with biopsy specimens helps pathologists be more consistent and accurate in diagnosing oral epithelial dysplasia.
Study design: Each of six board-certified oral and maxillofacial pathologists examined the same set of 120 oral biopsies (involving diagnoses ranging from hyperkeratosis to severe epithelial dysplasia); they had examined these same biopsies in a previous study, but this time the clinical information was provided for each case. The examiner's diagnosis was compared to the sign-out diagnosis for each case.
Results: Rates of exact agreement with the sign-out diagnosis averaged 38.5%, and there was 85.4% agreement within one histologic grade. The rate of agreement in distinguishing epithelial dysplasia from no dysplasia was 71.4%. These results, when compared to those from a previous study in which the same examiners had evaluated the same slides but without clinical histories, represent a 2.5% to 20% decrease for exact agreement among the six pathologists, a 0% to 8.5% decrease for agreement within one histologic grade, and a 0% to 23.4% decrease for agreement regarding the presence or absence of epithelial dysplasia.
Conclusions: When clinical information was used, accuracy and consistency among board-certified oral and maxillofacial pathologists in the diagnosis of oral epithelial dysplasia was not improved. In fact, there was a decrease in accuracy.