Aims and background: The reproducibility of histologic diagnoses, and in particular of the distinction between basal and squamous cell forms, has been evaluated in the framework of a multicenter case-control study (in Italy, France, Spain and Switzerland) aimed to assess the causes of non-melanocytic skin cancers.
Methods: A panel composed of 10 pathologists from the collaborating centers was appointed. A total of 1,774 slides of routine diagnoses were blindly reviewed by a second panelist; discordant diagnoses underwent a third examination. Controversial and interesting cases were discussed during general sessions.
Results: The results showed a high degree of concordance (99.5%), with a Cohen's kappa of 0.85 (95% CI, 0.77-0.94) in the assessment of malignancy of lesions. The concordance in the distinction between large morphologic groups, basal cell carcinoma and squamous cell carcinoma was also high (Cohen's kappa = 0.85; 95% CI, 0.82-0.89). The assessment of histologic subtypes, degree of invasion and differentiation showed a lower degree of concordance, presumably as a consequence of a weaker consistency in the relevant classifications.
Conclusions: The reproducibility study has therefore validated the epidemiologic study and in particular allowed to correct some misclassifications that could have lessened the analysis of the case-control study. In general, because of its characteristics (number of pathologists and variety of their origins, the large number of cases examined, blind examination of histologic slides), the conclusions of the study may show a certain degree of generalization, at least with regard to the countries represented. Routine histologic diagnoses of cutaneous carcinoma showed a high degree of reliability with reference to the assessment of malignancy and the distinction between basal and squamous cell carcinoma morphotypes.