Establishment of a predictive nomogram for differentiated thyroid cancer: an inpatient-based retrospective study.

Introduction: Differentiated thyroid cancer (DTC) is the most common malignant tumour of the endocrine system. The aim of this study was to establish a nomogram for simply and effectively predicting DTC.

Material and methods: 464 inpatients who underwent thyroid nodule surgery were retrospectively analysed. Univariate logistic regression and multivariate logistic regression were used to analyse the risk factors of DTC. A nomogram was constructed for predicting DTC.

Results: In this study, multivariate logistic regression found that female sex, age < 55 years, solid composition, hypoechogenicity, irregular margin, microcalcification, taller-than-wide, and cervical lymphadenopathy were independent risk factors for DTC. The area the curve (AUC) of the nomogram model indicated an excellent predictive performance of 0.920 [95% confidence interval (CI): 0.888-0.952]. The best threshold for predicting DTC was 52.4%, with sensitivity and specificity of 91.9% and 81.0%, respectively.

Conclusions: we provided a simple, noninvasive, and accurate model for clinicians to predict DTC.