Digital subtraction shoulder arthrography in determining site and size of rotator cuff tear.

RATIONALE AND OBJECTIVES: To evaluate the value of digital subtraction arthrographic techniques in assessing the precise size and site of rotator cuff tear. PATIENTS AND METHODS: Prospectively interpreted findings of 29 digital arthrographs were compared with results after surgery in 22 patients and with results after arthroscopy in 7 patients. RESULTS: Surgery or arthroscopy revealed 14 full-thickness and 2 partial-thickness tears. Digital subtraction arthrography showed 88% (14 of 16) of the rotator cuff tears, 86% (12 of 14) of the full-thickness tears, and 100% (2 of 2) of the partial-thickness tears. The site of a tear could be shown in 86% (12 of 14), but the size was evaluated to be less extensive than that seen at surgery in most tears (63%, 10 of 16). CONCLUSIONS: The site of tears could be demonstrated in supraspinatus tendon tears, but measurements of the size of the tears were accurate only in small full-thickness and in partial-thickness tears. Very anterior and posterior tears may not be diagnosed with this method, and postexercise films were necessary to avoid false-negative findings.