Premium
Is the reproducibility of shear wave elastography of thyroid nodules high enough for clinical use? A methodological study
Author(s) -
Swan Kristine Zøylner,
Nielsen Viveque Egsgaard,
Bibby Bo Martin,
Bonnema Steen Joop
Publication year - 2017
Publication title -
clinical endocrinology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.055
H-Index - 147
eISSN - 1365-2265
pISSN - 0300-0664
DOI - 10.1111/cen.13295
Subject(s) - thyroid nodules , reproducibility , medicine , elastography , thyroid , inter rater reliability , nodule (geology) , nuclear medicine , cutoff , radiology , ultrasound , rating scale , mathematics , statistics , paleontology , biology , physics , quantum mechanics
Summary Objective To systematically assess the reproducibility of thyroid ultrasonographic shear wave elastography ( SWE ). Context SWE has been suggested as a potential tool for thyroid nodule evaluation, but assessment of its reproducibility has been insufficiently addressed. Design SWE examinations were performed prospectively by two investigators. Patients Seventy‐two patients (male/female: 19/53; mean age: 53 ± 14 years; malignant/benign 17/55) undergoing thyroid surgery were enrolled in the study. Measurements Repeated and blinded measurements of elasticity index ( EI ) in predefined regions of interest ( ROI ) were collected. The inter‐ and intrarater agreement, along with the day‐to‐day agreement, was evaluated in terms of the 95% limits of agreement ( LOA ). Results are presented as a ratio, by which 1·0 indicates perfect agreement. Results The interrater, intrarater and day‐to‐day LOA showed ratios between repeated measurements of 1·7–3·6, 1·8–3·7 and 2·2–2·9, respectively. These values reflect a low to moderate degree of agreement for all EI outcomes. The interrater LOA was higher for malignant nodules compared with benign nodules for six of seven EI outcomes ( P < 0·001–0·03). The proportion of agreement calculated from the optimum cutoff point for differentiating malignant from benign nodules was 63–88% for the investigated EI outcomes. Conclusions In this methodological study, EI measured by thyroid SWE seems suboptimal for clinical use, due to a low inter‐ and intrarater agreement. That EI varies from day to day furthermore jeopardizes the validity of the method. Although the proportion of agreement was acceptable for some EI parameters, it is questionable whether EI assessments can reliably differentiate malignant from benign nodules in the individual patient.