Comparison of the diagnostic accuracy of thyroid fine-needle aspiration in follicular-patterned lesions using a 5-tiered and a 6-tiered diagnostic system: a double-blind study of 140 cases with histological confirmation.

Five-tiered and 6-tiered systems for reporting thyroid fine-needle aspiration (FNA) results are used widely throughout the world. In this study, we present a double-blind study of histologically confirmed follicular-patterned neoplasms and evaluate the cytological classification of the same lesions according to both systems. One hundred and forty consecutive surgically resected thyroid follicular-patterned lesions with a diagnostic preoperative FNA were retrieved from our archive. Two cytopathologists, who were blinded to all clinical information, classified each FNA case according to their respective routine diagnostic reporting system (5-tiered or 6-tiered). Interobserver variability was assessed using Cohen's Kappa (K) coefficient. Diagnostic accuracy was determined by measuring sensitivity and specificity. Receiver operator characteristic (ROC) curves were calculated for each cytopathologist. The 140 thyroid FNAs included histologically confirmed nodular hyperplasia, follicular adenomas, follicular carcinomas, and papillary carcinomas, follicular variant (35 cases for each) obtained from 104 females and 36 males with a mean age of 48.8 years and a mean tumor diameter of 27.8 mm. Negative predictive values (PV) for benign cases were 72.2% and 68.8% in the 5-tiered and 6-tiered systems, respectively (P = 0.7009). Positive PV were 100% for malignant cases in both systems. The sensitivity (78.6% vs. 72.9%, P = 0.4305), specificity (55.7% vs. 47.1%, P = 0.3103), and diagnostic accuracy (67.1% vs. 60.0%, P =0.2143) were similar between the systems. ROC curves almost entirely overlapped (P = 0.8937). Both the 5-tiered and 6-tiered systems show similar diagnostic accuracy in follicular-patterned lesions, further supporting the adoption of a common reporting system for thyroid cytopathology.