In the framework of the quality assurance of the pathologists involved
in the screening campaign it was proposed to test the reproducibility
of the classical SCARFF, BLOOM and RICHARDSON (SBR) grade and the ELS
TON-ELLIS grade (E-E) mentioned in the recent European recommendations
Twenty-four cases of invasive carcinoma were analyzed by 21 pathologi
sts. Sixty-eight per cent agreement was obtained for the SBR grant and
69 % for the E-E grade in the first round and vice versa in the secon
d round. The mean value of the kappa test,vas 0.54 with SBR and 0.53 w
ith E-E. The differentiation of the carcinoma was found to be the most
reliable index (kappa scope greater than 0.5), followed by the mitoti
c index (kappa score = 0.39 for SBR and 0.36 for E-E), and the nuclear
polymorphism obtained the lowest score (kappa score ranging from 0.33
to 0.38). In our hands, the use of the E-E grade did not improve the
reproducibility particularly for the mitotic index. Our results sugges
t however that the reproducibility may be improved by training.