Background: The 2015 American Thyroid Association (ATA) guidelines recommend the following size cut-offs based on sonographic appearances for subjecting nodules to fine-needle aspiration (FNA) biopsy: low-risk 15mm, intermediate-risk and high-risk 10mm.
Objective: We conducted a ‘real world’ study evaluating the diagnostic performance of the ATA cut-offs against increased thresholds, in the interest of safely limiting FNAs.
Methods: We performed a retrospective analysis of prospectively collected data on 604 nodules which were sonographically risk stratified as per the ATA guidelines and subsequently subjected to ultrasound-guided FNA. Nodules were cytologically stratified into ‘benign’ (Bethesda class 2) and ‘non-benign’ (Bethesda classes 3-6). We obtained the negative predictive value (NPV), accuracy, FNAs that could be spared, missed ‘non-benign’ cytologies and missed carcinomas on histology, according to the ATA cut-offs compared to higher cut-offs.
Results: In low-risk nodules, the high performance of NPV (≈91%) is unaffected by increasing the cut-off to 25mm, and accuracy improves by 39.4%; 46.8% of FNAs could be spared at the expense of few missed B3-B6 cytologies (7.9%) and no missed carcinomas. In intermediate-risk nodules, a 15mm cut-off increases the NPV by 11.3% and accuracy by 40.7%. The spared FNAs approach 50%, while B3-B6 cytologies are minimal, with no missed carcinomas. In high-risk nodules, low NPV (<35%) and accuracy (<46%) were obtained regardless of cut-off. Moreover, the spared FNAs achieved at higher cut-offs involved numerous missed ‘non-benign’ cytologies and carcinomas.
Conclusion: It would be clinically safe to increase the ATA cut-offs for FNA in low-risk nodules to 25mm and in intermediate-risk nodules to 15mm.