Objective
To determine the diagnostic value of ultrasonography in detection of the partial and full-thickness rotator cuff tear compared with arthrography. Method: One hundred twenty one cases were diagnosed as the patients with rotator cuff tear. Patients with full-thickness tear were divided into subgroups based on the size of small (＜1 cm), large (1∼3 cm) and massive (＞3 cm) tear. With arthrography as the standard of comparison for diagnosis of rotator cuff tear, full-thickness tear was confirmed by leakage of contrast extending into the subacromial-subdeltoid bursa. Results: Ultrasonography detected full-thickness tear in 90 cases, partial-thickness tear in 24 cases and no tear in 7 cases. A comparison of the results from ultrasonography and arthrography demonstrated a sensitivity of 83.8%, a specificity of 43.9% and an accuracy of 70.2% for detection of full-thickness tear. The size of supraspinatus full-thickness tear in inconsistent group of ultrasonographic and arthrographic findings was significantly smaller than that of consistent group. Conclusion: Ultrasonography may be used as the primary screening test for patients who may have suspected rotator cuff tears. However, other diagnostic techniques are recommended to confirm in patients who show small full-thickness tear on ultrasonography. (J Korean Acad Rehab Med 2006; 30: 362-367)
