Sonographically detected subacromial subdeltoid bursal effusion and bicepstendon sheath fluid: Reliable signs of rotator cuff tear?

Purpose. Our purpose was to determine the association between sonographical ly detected subacromial/subdeltoid (SA/SD) bursal and biceps tendon sheath effusions and arthrographically proven rotator cuff tears. Methods. Shoulder sonography reports and sonograms of 105 shoulders in 102 patients who also underwent arthrography were retrospectively reviewed for the presence of fluid within the biceps tendon sheath and SA/SD bursa. Repo rts and sonograms for 151 asymptomatic shoulders were also reviewed. Results. Biceps tendon sheath effusion and/or bursal fluid were detected in 50 (48%) of 105 shoulders. Fifty-one patients had rotator cuff tears; 28 o f them had effusions at 1 or both sites. The sensitivity, specificity, and positive predictive value (PPV) of biceps tendon sheath effusions for diagn osing rotator cuff tear were 35%, 74%, and 56%, respectively. For SA/SD bur sal effusions, the sensitivity, specificity, and PPV were 8%, 94%, and 57%, respectively. For combined biceps tendon sheath and bursal effusions, the sensitivity, specificity, and PPV were 12%, 91%, and 54%, respectively. The re was no statistically significant association between rotator cuff tears and effusions in the biceps tendon sheath, SA/SD bursa, or both. Among the 151 asymptomatic shoulders, 12 (7.9%) had biceps tendon sheath fluid, 5 (3. 3%) had SA/SD bursal effusion, and 2 (1.3%) had both biceps tendon sheath a nd bursal effusions. Conclusions. The sonographic detection of intraarticular fluid, SA/SD bursa l fluid, or both has a low sensitivity and PPV in the diagnosis of rotator cuff tears. Isolated intra-articular and/or SA/SD bursal effusions are not reliable signs of rotator cuff tear. (C) 1999 John Wiley & Sons, Inc.