Diagnostic accuracy and ability to reduce unnecessary FNAC: A comparison between four Thyroid Imaging Reporting Data System (TI-RADS) versions

Abstract

BackgroundThyroid Imaging Reporting Data System (TI-RADS) is used to characterize thyroid nodules while reducing unnecessary FNAC. Over the years, several versions of TI-RADS have been developed but there is no consensus on which TI-RADS is the best system. This study aimed to compare the diagnostic accuracy and ability of ACR TI-RADS, EU TI-RADS, K TI-RADS, AI TI-RADS to eliminate unnecessary FNAC. MethodsIn this prospective study, thyroid nodules were characterized by using the four TI-RADS systems and US-guided FNAC was done for nodule with the highest ACR TI-RADS score. Correlation between TI-RADS and FNAC results were analyzed. ResultsOut of 244 thyroid nodules, 100 nodules with either size <1 cm (43 nodules) non-diagnostic or inconclusive FNAC results (57 nodules) were excluded. Seven nodules (4.9%) were confirmed to be malignant on FNAC. K TI-RADS showed 100% sensitivity and NPV but the lowest specificity (40.2%). EU TI-RADS had the highest specificity (83.2%) but the lowest sensitivity (57.1%) and NPV (97.4%). ACR TI-RADS had an average sensitivity (85.7%) and NPV (98.6%). The specificity of ACR TI-RADS (51.1%) was lower than EU TI-RADS but higher than K TI-RADS. AI TI-RADS showed higher specificity (61.8% vs 51.1%, p < 0.05) but comparable NPV and sensitivity to ACR TI-RADS. AI TI-RADS was able to avoid the highest number of unnecessary FNAC (62.5%) followed by ACR TI-RADS(54.2%), EU TI-RADS(37.5%) and K TI-RADS(11.8%). ConclusionAI TI-RADS is a more simple scoring system with better overall diagnostic performance and ability to exclude unnecessary FNAC with high NPV. Advances in knowledgeHighest number of unnecessary FNAC thyroid could be prevented by applying AI TI-RADS.