Abstract
INTRODUCTION/PURPOSE: One examiner might not be able to acquire and interpret every image when examining rotator cuff tears in large cohort studies due to logistical issues. The aim of this study was to determine the repeatability of supraspinatus tear size measurements performed by multiple observers using previously acquired ultrasound images by a separate set of examiners. METHODS: Seventy-two ultrasound examinations of individuals with symptomatic isolated supraspinatus tears were performed by five ultrasound examiners. Fifteen examinations were randomly chosen to be evaluated by three observers with varying amounts of ultrasound experience using repeated measurements. The maximum anterior-posterior tear size using these previously acquired ultrasound examinations was assessed and the repeatability and minimal detectable change were calculated. RESULTS: The intra-observer repeatability was improved with greater experience to 0.87 (95% confidence interval [CI] 0.72-0.95) but was 0.51 (95% CI 0.20-0.77) and 0.75 (95% CI 0.51-0.90) for observers with low experience. For all observers, the inter-observer repeatability was 0.32 (95% CI 0.00-0.65). For all observers, the minimal detectable change was 4.6 mm and was lower in observers with more experience. DISCUSSION: Changes in supraspinatus tear size of 4.6 mm can be detected using previously acquired ultrasound images. This is comparable to ultrasound examinations performed and measured by the same individual as previously reported in the literature. Repeatability is poor among examiners with less experience. CONCLUSION: Clinical studies that cannot utilise measurements by a single examiner can be performed within current thresholds for determining changes in tear size of the supraspinatus tendon. LEVEL OF EVIDENCE: III.