112 - Predictive Value of TI-RADS Scores for Thyroidectomy and Assessment of Inter-Rater Variability

Purpose: To elucidate the relationship between the total score and individual components of the TI-RADS thyroid nodule scoring system, the Bethesda grade following FNA biopsy, the need for thyroidectomy, and likelihood of malignancy on final surgical pathology. Interobserver variability among radiologists and pathologists will also be analyzed.

Materials and Methods: Retrospective chart review was conducted of patients at a single institution over a 5-year period who received thyroid nodule FNA biopsies and had a preceding ultrasound report with a TI-RADS score of the corresponding nodule. Spearman’s correlation was calculated to determine if there was a correlation between TI-RADS scoring and Bethesda grading. Logistic regression evaluated whether TI-RADS score predicted thyroidectomy and whether individual features, including size, predicted benign or malignant pathology. The Kruskal-Wallis test was conducted to compare TI-RADS scores assigned by different radiologists and Bethesda scores assigned by different pathologists. Statistical analysis was conducted using IBM SPSS Statistics.

Results: Data from 220 nodules were collected, with 177 of the nodules occurring in female patients (80.5%). Spearman’s correlation demonstrated a weak but positive association between TI-RADS and Bethesda scores (ρ = 0.206, p = 0.002). Logistic regression showed that the TI-RADS score was an independent predictor of thyroidectomy, with a 1.9-fold increase in odds of undergoing thyroidectomy (OR = 1.92 [1.13-3.26], p = 0.016) and finding subsequent malignancy (OR = 1.91 [1.02-3.57], p = 0.042) for each point increase in the TI-RADS score. Among the TI-RADS features, microcalcifications were associated with greater odds of malignancy (OR = 8.13 [1.52-43.6], p = 0.014). The Kruskal-Wallis test identified a significant difference between radiologists (H(19) = 33.94, p = 0.010), suggesting that there is inter-rater variability in TI-RADS scoring. However, no statistically significant inter-rater variability in Bethesda scores was noted between pathologists (H(11) = 8.82, p = 0.639).

Conclusion: Higher TI-RADS scores are indeed associated with a greater likelihood of thyroidectomy following FNA biopsy, with microcalcifications being a predictive feature of malignancy. However, significant inter-rater variability among radiologists highlights the need for standardization of grading TI-RADS components on ultrasound reports.