Interobserver Agreement in the Diagnosis of Anal Dysplasia: Comparison Between Gastrointestinal and Gynaecologic Pathologists and Utility of Consensus Conferences.

INTRODUCTION Management of anal dysplasia relies on the accurate diagnosis of anal biopsy specimens. As institutions move toward subspecialty signout (SSSO), decisions must be made regarding whether to assign anal biopsies to the gastrointestinal (GI) or gynaecologic (GYN) pathology service. MATERIALS AND METHODS We identified 200 archival tissue biopsies of anal mucosa and circulated them among three GI pathologists and three GYN pathologists. Each pathologist separately scored each biopsy as normal, atypical, LSIL, or HSIL. Every case that was called HSIL by at least one pathologist was stained with p16 immunostain and a "gold standard" interpretation of whether a case represented HSIL was made. RESULTS The GI pathologists agreed on 97 (49%) cases prior to consensus; the GYN pathologists agreed on 33 (17%). The sensitivities of the 3 GI pathologists in detecting HSIL against the "gold standard" were 47%, 100%, and 21%, and for the GYN pathologists the sensitivities were 74%, 89%, and 84%; the sensitivities of both the GI and GYN consensus diagnoses were 74% each. The specificities of the 3 GI pathologists in detecting HSIL were 99%, 90%, and 100%, and for the GYN pathologists the specificities were 99%, 92%, and 91%; the specificities of both the GI and GYN consensus diagnoses were 100%. CONCLUSIONS A mild to moderate degree of interobserver variability exists in the diagnosis of anal dysplasia among pathologists. Our study does indicate the utility of some form of consensus conference, as overall agreement among GI pathologists and among GYN pathologists improved following in-person consensus.