Background Thyroid cancer is the most common malignancy of the head and neck, resulting in a high volume of referrals for evaluation of thyroid nodules. Ultrasound risk stratification using the Thyroid Imaging Reporting and Data System (TI-RADS) and cytological assessment via the Bethesda System for Reporting Thyroid Cytopathology are central to preoperative decision-making. However, their diagnostic performance varies when validated against final histopathology. Objectives The primary objective of this study was to assess the concordance between TI-RADS, Bethesda cytology, and final histopathology in surgically managed patients. The secondary objective was to evaluate preoperative malignancy detection rates and identify missed malignancies or discrepant cases. Methods This retrospective observational study included patients who underwent total or hemithyroidectomy at Salmaniya Medical Complex, Manama, Bahrain, between 2017 and 2024. Preoperative TI-RADS and Bethesda classifications, demographic data, and nodule characteristics were analyzed. Diagnostic performance was assessed for both modalities and further stratified by histological subtype, indeterminate categories (TI-RADS 3-4, Bethesda III-IV), tumor size, age, and sex. Results Of 594 patients who underwent thyroid surgery, 159 cases with histologically confirmed malignancy and complete preoperative data were included in the final analysis (mean age: 44.7 years; females: 127 (79.9%)). Ultrasound correctly identified 137 (86.2%) of malignant nodules, while fine needle aspiration cytology (FNAC) detected 134 (84.3%). Both modalities demonstrated reduced detection rates for follicular carcinoma, with ultrasound identifying seven of 10 cases (70.0%) and cytology identifying six of 10 cases (60.0%). Malignancy was observed within indeterminate and low-risk categories, including TI-RADS 3 nodules and Bethesda III cytology. Among Bethesda III nodules (n= 52), 15 cases (28.8%) were malignant on final histopathology. Additionally, a substantial proportion of benign nodules classified as high-risk on ultrasound (TI-RADS ≥4) exhibited benign or indeterminate cytology. Larger tumors (>4 cm) were more reliably detected, whereas age, sex, and tumor location did not significantly affect diagnostic performance. Conclusion Ultrasound TI-RADS and Bethesda cytology demonstrate high but imperfect sensitivity for the preoperative detection of thyroid malignancy, with ultrasound showing slightly superior performance. Diagnostic limitations were most evident in follicular-pattern tumors and indeterminate nodules. These findings emphasize the importance of integrated ultrasound and cytological assessment to improve risk stratification and guide surgical decision-making in patients with thyroid nodules.
Alboosta et al. (Tue,) studied this question.