Thyroid ultrasound and nodule malignancy risk: a "real world" assessment of ultrasound reporting and agreement of ultrasound-based malignancy risk estimates with cytology and histology findings.

Aims: Thyroid nodule malignancy risk is increasingly estimated using ultrasound characteristics. We assessed ultrasound reports of nodules and compared ultrasound-based malignancy risk assessments with cytology and histology findings.

Methods: We identified patients with thyroid ultrasound (55% by private provider, 45% by DHB) and cytology at CMDHB over 18 months. Malignancy risk for each nodule was categorised based on the ultrasound report, then using ultrasound images with the local CMDHB approach and American Thyroid Association guidelines, and then was compared with cytology/histology results.

Results: 36/91 nodules (84 patients) had abnormal (Bethesda 3-6) cytology. Forty-eight patients (54 nodules) underwent thyroid surgery and 13 nodules (12 patients) had thyroid cancers. Most ultrasound reports did not mention nodule malignancy risk characteristics (range 13-98%) or a malignancy risk estimate (66/91). 12/33 nodules with benign (Bethesda 2) cytology and 18/36 nodules with abnormal (Bethesda 3-6) cytology were considered intermediate/high risk of malignancy by ultrasound; none and seven, respectively, had cancer identified subsequently. In 18 nodules considered low risk by ultrasound, four cancers were identified.

Conclusions: Most ultrasound reports contained insufficient information about nodule malignancy risk to allow an independent assessment. Agreement between cytological/histological findings and malignancy risk estimates using ultrasound was not high.