Burg G, Zwingers T, Staegemeir E, Santucci M
Department of Dermatology, University Hospital, Zürich, Switzerland.
Dermatol Clin. 1994 Apr;12(2):311-4.
Histologic and cytologic evaluation of CTCL is one of the most difficult challenges in dermatopathology. It was the purpose of this study to evaluate the reliability and reproducibility of histologic criteria in CTCL and the interrater and intrarater variabilities. An overall of 2375 ratings of sections from 41 patients with CTCL were given by a panel of 22 dermatopathologists and pathologists familiar with lymphoma histology, some of whom were looking repeatedly at the same series of sections. The results clearly indicate that interrater as well as intrarater variabilities are significant and can be decreased by discussion preceding the evaluation as well as by repeated rating indicating a learning effect. Therefore, histologic evaluations of criteria have to be taken with caution, and statistical evaluation of level of agreement expressed by kappa coefficient is always mandatory in this type of study.
皮肤T细胞淋巴瘤（CTCL）的组织学和细胞学评估是皮肤病理学中最具挑战性的难题之一。本研究旨在评估CTCL组织学标准的可靠性、可重复性以及不同评级者之间和同一评级者内部的变异性。由22名熟悉淋巴瘤组织学的皮肤病理学家和病理学家组成的小组，对41例CTCL患者的切片进行了总计2375次评级，其中一些人反复查看同一组切片。结果清楚地表明，不同评级者之间以及同一评级者内部的变异性都很显著，通过评估前的讨论以及重复评级（显示出学习效应）可以降低变异性。因此，对标准的组织学评估必须谨慎进行，在这类研究中，用kappa系数表示的一致性水平的统计评估始终是必要的。
