Objective: Fine needle aspiration cytology is the first line of investigation for thyroid lesions. Despite standard reporting formats, the diagnostic accuracy varies across institutions. In this study, we have reviewed our discordant cases on cytology and histopathology and analyzed the diagnostic errors.

Material and Methods: The thyroid cases with discrepant cytology and histopathology reports for a period of five years were analyzed for diagnostic errors. The papillary thyroid carcinoma (PTC) cases were studied in detail for all diagnostic parameters. Nuclear scoring was used to improve the detection of PTC.

Results: Of the 166 cases, 18 (10%) had discrepant diagnoses. The sensitivity was 65.62% (CI 46.81-81.43%), specificity 94.78%, positive predictive value 75%, negative predictive value 92.03%, positive likelihood ratio 12.56, negative likelihood ratio 0.36, false positive rate 5.2%, false negative rate 34.3% and accuracy 89.16%. False negative (malignant cases diagnosed as benign) was due to inadequate/wrong site sampling, benign clusters/ cyst macrophages, marginal flares, thin colloid, larger fragments of calcification, and subtle nuclear features. An interesting flower head-like structure was observed in PTC cases. Nuclear scoring on false negative cases improved our diagnostic accuracy. False positivity was due to vigorous aspiration and over-interpretation of nuclear features.

Conclusion: Analysis of our discrepant cases highlighted the importance of multiple passes, sampling all nodules, and ultrasound-guided aspiration to reduce sampling error. Application of nuclear scoring reduced overdiagnosis and missing out on PTC. Tissue fragments and hypercellularity were the major misleading factors in false positive cases.