[Chronic shoulder pain. Ultrasonography versus surgery].

This work was aimed at investigating US reliability in the diagnosis of rotator cuff lesions. A series of 554 patients with a chronically painful shoulder submitted to US was retrospectively reviewed to search for any possible rotator cuff lesion. US demonstrated total or partial cuff tears in 82 potentially surgical patients (14.8%) Then, in this group, the occurrence of every US sign was investigated. Cuff disappearance was observed in 42 patients (51.2%), convexity loss in 20 (24.4%), cuff thinning in 17 (20.75%) and focal discontinuity in 3 patients only (3.7%). In the second part of the study, we considered only 33 surgical patients. Surgery confirmed cuff tears in 30 patients (90.9%), whereas in 3 false-positive patients US had overestimated focal hyperechoic areas and tendon thinning. Thus, our study confirms the almost absolute sensitivity of some US signs of cuff lesions, e.g., cuff disappearance and convexity loss, in the screening of the patients with a chronically painful shoulder. On the contrary, when focal hyperechoic lesions or tendon thinning are present and US poorly demonstrates them, potentially surgical patients must be submitted to MRI.