Interrater Reliability of Various Thyroid Imaging Reporting and Data System (TIRADS) Classifications for...

Area of Science:

• Radiology
• Oncology
• Endocrinology

Background:

• Thyroid ultrasound (US) is a primary diagnostic tool for thyroid disease management.
• US assessment is operator-dependent, with limited data on interrater variability.
• Evaluating interrater reliability is crucial for standardizing US assessments of thyroid nodules.

Purpose of the Study:

• To assess the interrater reliability of thyroid nodule ultrasound assessments.
• To estimate the diagnostic accuracy of various Thyroid Imaging Reporting and Data Systems (TIRADS).
• To compare the performance of ACR-TIRADS, Siriraj-TIRADS, and EU-TIRADS systems.

Main Methods:

• Retrospective review of 24 malignant and 84 benign thyroid nodules.
• Two blinded observers independently assessed US images using ACR-TR, Siriraj-TR, and EU-TR guidelines.
• Interrater reliability calculated using Cohen’s Kappa statistics; diagnostic accuracy metrics computed.

Main Results:

• Substantial agreement for composition (K=0.616), fair for echogenicity (K=0.327) and echogenic foci (K=0.288), slight for margin (K=0.143).
• Moderate agreement for ACR-TIRADS (K=0.500), fair for EU-TIRADS (K=0.209), slight for Siriraj-TIRADS (K=0.114).
• ACR-TIRADS showed highest sensitivity (75-79.2%) and NPV (89.1-90.4%) for malignancy; Siriraj-TIRADS had higher specificity (84.5-89.3%).

Conclusions:

• The ACR-TIRADS system exhibited the highest interobserver agreement and a trend towards higher sensitivity and NPV for diagnosing malignant thyroid nodules.
• The Siriraj-TIRADS system demonstrated higher specificity and accuracy but lower interobserver agreement.
• Standardization of TIRADS application is essential to improve diagnostic consistency in thyroid nodule assessment.