This is a retrospective study of 33 of 48 arthroscopic repairs of partial rotator cuff tears performed more than 2 years ago.

Repairs were done by one of four techniques: transtendon, side-to-side, completion of tear and side-to-side, or completion of tear and anchor repair. All patients completed UCLA shoulder score and Western Ontario Rotator Cuff (WORC)© score questionnaires.

We had 91% excellent or good UCLA shoulder scores and 85% excellent or good WORC© scores. Repair technique, tear classification, comorbidity or postoperative stiffness (experienced by 36% of patients) did not significantly affect outcome scores.

This study was designed to evaluate the results of arthroscopic rotator cuff repair at a minimum follow-up of 2 years.

Only isolated full-thickness rotator cuff tears were included in the study. Of 63 cases that met the criteria, 51 were followed up. Results were measured with pre-operative and postoperative UCLA shoulder scores, Western Ontario Rotator Cuff (WORC)© scores, range of motion, strength and radiographs. The time to recovery and return to work, complications and patient satisfaction were also recorded. Data from various tear sizes were analysed to determine significant differences.

At the most recent follow-up, all patients had less pain and better function. Patients rated 48 shoulders (94%) satisfactory. Mean forward flexion was 170.4° and mean manual strength was 4.8/5. Significant strength differences in flexion and external rotation were found between various sizes of tear (p < 0.01). Mean UCLA scores for all tear sizes significantly improved from a preoperative 10.3 (±2.4) to a postoperative 32.1 (±4.3). The mean for small tears was 35, for medium tears 33.3 and for large tears 30. This difference was statistically significant (p < 0.05). Excellent postoperative UCLA scores were achieved in 26 shoulders (51%) and good results in 17 (33%), with seven shoulders (14%) fair and one (2%) poor. The mean overall WORC© score was 86.8% (±17.1) of normal. The mean time to recovery was 5.1 months.

Arthroscopic rotator cuff repair has good results and can be done on an outpatient basis with few complications. The results appear to depend on the tear size.

In diagnosis and repair of partial subscapularis tendon tears, we used a simplified arthroscopic direct technique.

We used the anterosuperior arthroscopic portal to visualise the subscapularis tendon insertion, while probing and repairing from the adjacent anterior portal. Three anatomical dissections were done to define the insertion of the subscapularis tendon at the lesser tuberosity. While viewing from the anterosuperior portal, we repaired the subscapularis tendon with one or two suture anchors, inserted into the lesser tuberosity from the anterior portal. Suture management was via the standard posterior portal. From the anterior portal, a tendon-penetrating grasping device passed the sutures through the displaced subscapularis tendon. The arthroscopic knots were tied from the anterior portal.

Associated with the first 10 subscapularis repairs were six complete and four partial thickness supraspinatus/infraspinatus tears. There were no isolated subscapularis tears. Three patients had associated biceps lesions.

Subscapularis tears are often associated with supraspinatus and infraspinatus tendon tears. Direct anterosuperior viewing and anterior probing enables the surgeon to see and repair ‘hidden’ tears.