Background: Ultrasound is nowadays a method of choice for thyroid volume assessment. However, its disadvantage is some inaccuracy, which is said to be higher in huge, especially substernally extended goitres.
Aims: The aim of the study was to compare the US and CT thyroid volumetric measurements: multi-observers (CT MO) and one-observer (CT OO) to CT planimetry results (CT Pl) in patients with large goitres.
Materials & methods: The study material comprised 70 thyroid imaging examinations obtained from 35 patients with nontoxic goitres, scanned twice before and after radioiodine treatment. Mean thyroid volume was 88·97 ± 60·21 ml. Thirty-three thyroid scans revealed the extension below the jugular notch (mean of 2·46 cm). Thyroid volume in US, CT MO and CT OO was estimated using the ellipsoid formula. CT Pl was established a reference method.
Results: The mean thyroid volume in CT Pl was 88·97 ml (median 80·73, range 11·81 to 315·97). US underestimates thyroid volume by 7·55 ml (7·7%) with a sufficient correlation (R(2) = 0·89) and precision (20·37). CT OO is the closest and CT MO the most distant from CT Pl, with US between them in thyroid volume estimation. The percentage US bias is constant through all range of thyroid volume. There is no difference for percentage bias between US and CT Pl for goitres with (8·67%), and without (6·70%) substernal part.
Conclusion: US examination is sufficient for epidemiological studies, radioiodine activity calculation and goitre size assessment in everyday medical practice. Neither initial size of the goitre nor its substernal extension affects US assessment precision.
© 2014 John Wiley & Sons Ltd.