Abstract
INTRODUCTION: Musculoskeletal diagnostic ultrasound is an expanding clinical tool and increasingly used at point of care in clinical practice. Criticism around the accuracy of this modality is often attributed to operator dependence. This study evaluated the accuracy of a sonographer's reporting against shoulder surgery. METHODS: A musculoskeletal sonographer performed same-day ultrasound scans prior to elective shoulder surgery procedures. The sonographer was blinded to clinical history and previous investigations. A brief physical examination of resisted external rotation and resisted abduction strength was performed prior to scanning. Scan results were reported using British Elbow and Shoulder Society (BESS) reporting criteria. The sonographer attended surgery to gain immediate visual feedback. Sonographic reports and magnetic resonance imaging (MRI) reports were graded (1= complete agreement; 2 = minor discrepancy unlikely to alter patient care; 3 = potentially significant discrepancy; 4 = definite, significant discrepancy) against surgical findings. The accuracy of sonographic reporting was correlated with patients' body mass index (BMI). RESULTS: Fifty-four cases were compared and correlations evaluated. Grade 1 (complete agreement) was reported 87.0% (n = 47), Grade 2 (minor discrepancy) 9.3% (n = 5) and Grade 3 (potentially significant discrepancy) 3.7% (n = 2). Sonographic reports closely correlated (Grade 1 or 2 scores) with surgical findings in 96% of cases. CONCLUSIONS: The musculoskeletal sonographic reporting had a high correlation with surgery findings. Ultrasound was more accurate than MRI in reporting full thickness rotator cuff tears. Clinical weakness in resisted external rotation or abduction did not correlate with imaging. Ultrasound was found to be unreliable in detecting rotator cuff tears in participants with a BMI ≥ 35.