Prostate cancer: interobserver analysis.

Abstract

Abstract Because it has good intra and interobserver reproducibility, Gleason`s score is widely used for the histopathological assessment of prostate carcinomas. This was a nationwide study assessing the interobserver agreement of uropathologists, general pathologists and pathology residents, concerning the use of Gleason`s score. Slides from 30 patients, with a diagnosis of adenocarcinoma confirmed by analysis of the prostatectomy specimen and immunohistochemical analysis of high-molecular weight cytokeratin, were randomly selected from the needle-guided prostate biopsy specimens of the Cito Cytopathology Laboratory, Juiz de Fora, MG, Brazil, during the period June 1998 through July 1999. Each slide was photographed at three magnification powers: 40, 100 and 200x. The photos from the 30 cases were made available on-line, for grading of the biopsy materials, in the website of the Pathology Department of the School of Medicine of the Juiz de Fora Federal University (www.patologia.ufjf.br), along with an invitation letter and demographic questionnaire, addressed to pathologists accessing the PATOCTO website (patocito@yahoogroups.com.br). Grading results were graphically plotted according to the percentages of correct diagnoses and against the demographic data. “True” diagnosis was considered that for which Gleason`s score reached higher consensus. Mean total agreement was 65.20%, with maximum agreement of 79.68% for the score considered to be the “true” diagnosis. This reproducibility index with identical score, with + or – 1 variation in the score considered to be the “true” diagnosis, reached 85%, consistent with data by Gleason himself and the literature. This experience presents an important on-line tool which may be used for continuing medical education. Key words: Prostatic neoplasms; observer variation; Biopsy, Needle; internet;