Inter-observer reproducibility of the Royal College system for reporting thyroid cytology: Experience of the Egyptian National Cancer Institute

Abstract

Purpose: To assess the inter-observer agreement between 3 cytopathologists of thyroid FNAC using Royal College of Pathology reporting system.

Patients and methods: The study is a retrospective one conducted on 204 cases retrieved from the archives of the Cytology Unit, Pathology Department, National Cancer Institute, Cairo University during the time period from January 2016 to December 2016. Cases were diagnosed separately by 3 cytopathologists using the Royal College of Pathology classification system (RCPath), where Thy1, nondiagnostic; Thy2, nonneoplastic; Thy3a, atypical, Thy3f, follicular lesion; Thy4, suspicious of malignancy; and Thy5, malignant. Kappa statistics were used where combination of the agreement between the 3 observers simultaneously was done.

Results: There was a good overall agreement between the three observers regarding all categories (kappa statistics was 0.679). Perfect agreement was reported for Thy5 category (κ = 0.874), good agreement was observed for Thy1 and Thy2 (κ = 0.784 and 0.719, respectively). For Thy3a, Thy 3f and Thy 4, a moderate agreement was reported (κ = 0.407, 0.446 and 0.453 respectively). Combination of surgical categories (Thy3f, Thy4, and Thy5) achieved a good agreement (κ = 0.701) as well as for non-surgical categories (Thy1, Thy2, and Thy3a) (κ = 0.712).

Conclusion: RCPath reporting system for thyroid FNAC is clinically applicable and can be used for differentiation between benign cases needing observation and follow up on one hand, and malignant cases requiring surgical intervention on the other. The least inter-observer agreement (moderate agreement) was detected for Thy3a, Thy3f and Thy4 categories, necessitating multidisciplinary team meeting before any clinical decision.