Reproducibility of grading systems in oral epithelial dysplasia

Objective: To assess inter and intra observer variability in grading oral epithelial dysplasia (OED) using Smith and Pindborg grading system, WHO classification system and Brothwell DJ et al. classification system. Study design: In the study 45 histological tissues of dysplasia, 15 each of mild, moderate and severe dysplasia were coded and blindly graded by three observers in three grading systems. Further on the same observers graded 15 slides again of the previous 45 for analyzing the reproducibility in the three grading systems. The individual significance of various indicators of dysplasia among various grades of dysplasia was also assessed. Result: Inter observer agreement was significantly higher in Brothwell system as compared to WHO and Smith and Pindborg system. Intra observer agreement was significantly higher in Smith and Pindborg system, but the predictability and the probability index was distributed over a larger range in this system. Each indicator of dysplasia was also found to be statistically significant (P<0.05) for grading dysplasia. Conclusion: The present study puts forth the inherent intricacies in the grading of oral premalignant lesions. Key words:Carcinoma, dysplasia, grading systems, reproducibility.