Inter-examiner variability in grading oral epithelial dysplasia may result in suboptimal management of oral potentially malignant disorders

Abstract

Oral epithelial dysplasia (OED) grading determines the management guidelines for oral potentially malignant disorders (OPMDs). The subjectivity of OED grading considerably impacts its reliability. Objective: This study aimed to assess the reproducibility of and variability in diagnosing and grading OED by oral and medical pathologists, using the conventional WHO 2005 classification. Material & methods: Five oral pathologists and one medical pathologist individually examined 200 hematoxylin and eosin-stained histological slides diagnosed as OED from oral pathology archives at the University of Peradeniya. The most experienced examiner’s diagnoses represented the standard for evaluating inter-examiner variability using the unweighted Cohen’s kappa coefficient. Results: OED grading among all oral pathologists revealed moderate agreement (kappa value, 0.42–0.50), whereas the medical pathologist showed poor agreement (kappa value, 0.034). The accepted OED diagnoses were mild, 33%; moderate, 24.9%; severe, 32.4%; and no dysplasia, 9.7%. However, 86.5% of the diagnoses by the medical pathologist were mild-no dysplasia. Diagnoses of moderate and severe dysplasia had lesser reproducibility than those of no dysplasia. Conclusions: OED grading was only moderately reproducible among oral pathologists and poorly reproducible with regard to the medical pathologist. A more reliable OED grading system is required to improve reproducibility for optimal OPMD management and assessment.