ISSN 1470-3947 (print) | ISSN 1479-6848 (online)

Endocrine Abstracts (2019) 63 P364 | DOI: 10.1530/endoabs.63.P364

Sensitivity and specificity of ultrasound features in predicting malignant thyroid nodules

Rym Belaid, Ibtissem Oueslati, Wafa Grira, Ons Rejeb, Meriem Yazidi, Fatma Chaker & Melika Chihaoui


Department of Endocrinology, La Rabta Hospital, Tunis, Tunisia.


Background: Ultrasound is valuable for identifying many malignant or potentially malignant thyroid nodules. The aim of this study was to evaluate the efficacy of ultrasonographic (US) features in predicting the malignancy of thyroid nodules in a group of Tunisian patients.

Methods: We conducted a retrospective descriptive study including 100 patients collected in the endocrinology department of La Rabta hospital. All these patients underwent thyroidectomy for one or more suspicious thyroid nodules on clinical, ultrasound or cytological criteria. For each patient histological result of the surgical specimen was collected. Univariate analyse was performed to calculate the sensitivity, specificity, negative and positive predictive values of each US feature.

Results: Of the 100 patients, 87% were female. The overall mean age was 49.9±17.4 years (range: 12–79 years). On histocytopathology, 20% of the nodules were malignant. Ultrasound predictors of malignancy were a nodule size < 20 mm (OR =3.6; P=0.009), the solid composition (OR=52.6; P< 0.001), the ill-defined margins (OR=3.4; P=0.014) and the central vascularity (OR=8.5; P=0.008). The sensitivity and specificity of ultrasound features in predicting malignancy were microcalcification 45% and 69%; hypoechogenecity 85% and 29%; ill-defined margin 50% and 23%; solid echostructure 40% and 99%; heterogenicity 90% and 3%, central vascularisation 25% and 96%. Each ultrasound feature had negative predictive value ranging from 50% to 88% in malignant nodules.

Conclusions: Our study shows that no single US sign was reliable in differentiating all benign from malignant thyroid nodules, but many US features help in predicting the benign or malignant nature of a given nodule. Thus, a predictive model for malignancy using a combination of clinical, biochemical, and radiological characteristics may support clinicians in reducing unnecessary invasive procedures in patients with thyroid nodules.

Article tools

My recent searches

No recent searches.