Reproducibility of three classification systems of ductal carcinoma in situ of the breast using a web-based survey

This study assessed the degree of diagnostic agreement among pathologists between three classification systems of ductal carcinoma in situ of the breast (DCIS). Thirteen pathologists received the same set of digitized images of microscopy of 43 DCIS cases and answered a questionnaire containing the criteria to compose the three classification systems studied: Holland, modified Lagios, and Van Nuys. A computer program was created, which organizes the information collected from each pathologist, supplying the histological grading of the cases within the three classification systems. The results were analyzed using percental agreement and the Kappa test. Diagnostic agreement for the three DCIS of the breast classification systems presented K values that varied from 0.27 to 0.37. Among the three classifications used, most agreement was for Van Nuys, showing a Kappa index of 0.37. These results matched the interobserver agreements, with Kappa indices varying from 0.13 to 0.64 for the Holland classification; 0.23 to 0.61 for the modified Lagios classification; and 0.23 to 0.74 for the Van Nuys classification. Pathologists specialized in breast pathology showed greater reproducibility for all the criteria evaluated. Comparing the three classification systems, diagnostic agreement and accuracy were rated higher for the classification of Van Nuys compared to modified Lagios and Holland.