To evaluate the reliability of the arthroscopic assessment of full-thicknes
s rotator cuff tears, 117 cases were prospectively investigated by imaging,
arthroscopy, and open surgery. The confidence of the surgeon, his accuracy
, and the surgeon-dependent character of arthroscopic assessment were evalu
ated in terms of the description of the main anatomic parameters. The surge
ons were confident and accurate in diagnosing a full-thickness tear of the
supraspinatus, but they underestimated its coronal and sagittal extent and
its reducibility to the greater tuberosity. Conversely, the technique appea
red very accurate in describing the rotator interval. Endoscopic assessment
was particularly operator-dependent in the anteroposterior analysis of the
tear. This study shows the Limits of endoscopic assessment of full-thickne
ss rotator cuff tears. It illustrates the need for an adequate arthroscopic
technique with a thorough knowledge of normal and pathological anatomy of
the rotator cuff.