To report our experience of massive rotator cuff tear and evaluate the final outcome and factors influencing the final results.

We have performed 25 repairs (24 patients) of the massive rotator cuff tear. The average age 58 years old, mean follow-up was 22 months. Acromioplasty alone was performed in 7 cases and acromioplasty with direct repair was performed in 18 cases. We statistically analyzed the results by Student's T-test.

Postoperative pain and UCLA score improved 2.2, 32 from 7.4, 14 respectively. The gain of 32 degrees in forward flexion, 8 degrees in external rotation could be obtained. Six (24%) was excellent, sixteen (64%) was good and three (12%) was poor and eigthy-four percent of our series was satisfied. Preoperative range of motion (P=0.040) , the degree of degeneration and the magnitude of tendon retraction (P=0.013) influenced the final results, but age (P=0.232) , duration and severity of pain (P=0.370) did not. There was a significant difference between acromioplasty only and acromioplasty with direct repair even though this discrepancy is very small (P=0.046) . Retear was revealed in 8 cases (44%) among repaired rotator cuff.

Acromioplasty by itself could get a pain relief and functional improvement and a patient satisfaction in the treatment of massive rotator cuff tear regardless a reparability, but a more favorable result could be expected when a torn rotator cuff was repaired if possible.