The proximity of neural structures to the coracoclavicular ligaments limits the amount of coracoid process that can be harvested. The purpose of this study of 100 dry human scapulae was to define the anatomic limitations.

We found the mean measurement of the horizontal arm of the coracoid process anterior to the conoid tubercle was 21.5 mm (SD 0.9 mm). In 10% of the scapulae, it was larger than 30 mm. In 66%, the posterior aspect of the conoid fused with the vertical ramus and the lateral lip of the suprascapular notch.

This amount of coracoid appears to be large enough to expand the glenoid vault, and to hold two AO small fragment screws. It can be safely harvested if the conoid ligament is respected. Partial sacrifice of the trapezoid ligament is unavoidable, but does not compromise coracoclavicular stability. If the coracoid osteotomy is extended medial to the conoid tubercle it encroaches on the vertical ramus of the coracoid and can damage the suprascapular nerve. Posterior advancement of the osteotomy can extend onto the anterosuperior glenoid.

We evaluated the clinical outcome of arthroscopic labroplasty in 56 patients treated for shoulder instability owing to ligamentous laxity.

In our technique, the antero-inferior labral capsular complex is detached and mobilised from the glenoid. It is advanced superiorly and plicated to create a new labrum, retensioning the capsule and decreasing the articular volume. Usually, a rotator interval plication is also added. Postoperatively, patients wear an adduction sling for three weeks, but movement is permitted within pain limits.

The mean time to follow-up, when patients were clinically reviewed and assessed on the Walch-Du Play score, was 26 months (12 to 74). No intra-operative complications or nerve injuries were encountered. There was a single failure with frank redislocation. The mean Walch-Du Play score was 88/100 (10 to 100).

Redundant capsule and a hypoplastic labrum are common in unstable shoulders owing to ligamentous laxity. The labroplasty creates a ‘bumper’ and addresses the excess of capsule. In our short-term experience, this arthroscopic technique is superior to the open capsular shift.

Between 1996 and 2001 we used a modification of the Latarjet procedure to treat 70 patients with bony insufficiency of the glenoid. Our modification involves detaching a long piece of coracoid and rotating it to match its concave inferior surface with the surface of the glenoid. The coracoid graft is placed extra-articularly and the capsule repaired with bone anchors to the edge of the glenoid.

Postoperatively no sling is applied and rehabilitation is started early.

At a mean of 24 months (9 to 72) patients were clinically reviewed and assessed on the Walch-Du Play score. The results were excellent in 68%, moderate in 6% and poor in 1%. There were no redislocations.

The results were most satisfactory in this group of patients, most of whom participated in contact sports, where soft tissue procedures (e.g., open and arthroscopic Bankarts) carry unacceptable failure rates.