Scapular notching is of concern in reverse shoulder arthroplasty and has been suggested as a cause of glenoid loosening. Our purpose was to analyze in a large series the characteristics and the consequences of the notch and then to enlighten the causes in order to seek some solutions to avoid it. 430 consecutive patients (457 shoulders) were treated by a reverse prosthesis for various etiologies between 1991 and 2003 and analyzed for this retrospective multicenter study. Adequate evaluation of the notch was available in 337 shoulders with a follow-up of 47 months (range, 24–120 months). The notch has been diagnosed in 62% cases at the last follow-up. Intermediate reviews show that the notch is already visible within the first postoperative year in 82% of these cases. Frequency and grade extension of the notch increase significantly with follow-up (p< 0.0001) but notch, when present, is not always evolutive. At this point of follow-up, scapular notch is not correlated with clinical outcome. There is a correlation with humeral radiolucent lines, particularly in metaphyseal zones (p=0.005) and with glenoid radiolucent lines around the fixation screws (p=0.006). Significant preoperative factors are: cuff tear arthropathy (p=0.0004), muscular fatty infiltration of infraspinatus (p=0.01), narrowing of acromio-humeral distance (p< 0.0001) and superior erosion of the glenoid (p=0.006). It was more frequent with superolateral approach than with deltopectoral approach (p< 0.0001) and with standard cup than with lateralized cup (p=0.02). We conclude that scapular notching is frequent, early and sometimes evolutive but not unavoidable. Preoperative superior glenoid erosion is significantly associated with a scapular notch, possibly due to the surgical tendency to position the baseplate with superior tilt and/or in high position which has been demonstrated to be an impingement factor. Preoperative radiographic planning and adapted glenoid preparation are of concern.

Purpose: Outome after arthroscopic management of anterior instability of the shoulder has varied since the early series. The results proposed at the 1993 symposium of the French Society of Arthroscopy suggest we should be using this technique with prudence. We report here our experience with patients operated on between 1993 and 1997 who were selected on the basis of the 1993 conclusions that excluded patients with multiple recurrent instability and fractures of the anterior rim of the glenoid cavity.

Material and methods: Sixty-seven shoulders were operated on between 1993 and 1997. Mean follow-up for 58 of these shoulders (86%) was five years. These 58 patients constituted the study group. There were 31 men and 27 women, mean age 25 years, who had 30 recurrent shoulder dislocations, 12 shoulder subdislocations and 16 painful unstable shoulders. Forty-six percent of the patients participated in competition-level sports with forced shoulder movements in 39.6% of the cases. The surgical technique involved retightening the inferior glenohumeral ligament that was fixed with resorbable sutures. Immobilisation with elbow-to-body contention was strictly applied for three weeks at least followed by progressive rehabilitation exercises until renewed sports activities starting four months postoperatively.

Results: The mean overall Duplay score was 85.5 (sport 21/25; stability 18/25; mobility 24/25; pain 22/25). Outcome was good and very good in 82.7% of the patients, fair in 8.6%, and poor in 5 (recurrence). Subjectively, 55% of the patients were very satisfied, 27.5% were satisfied, 15.5% were disappointed and 1.7% were displeased. There were four complications (one infection cured with antibiotic therapy with a final score of 100 and three serious cases of capsulitis that recovered before one year). Gender, age, type and duration of instability, level of sports activity, and articular laxity appeared to affect outcome.

Discussion: The rate of failure (8.6%) is similar to that with open surgery (4.6% in the SOFCOT symposium 1999) and would be well below the rates observed in the 1993 arthroscopy series although the different patient selection does not allow valid comparison.

Conclusion: Arthroscopic stabilisation of the shoulder is a technically difficult procedure that has progressively shown its effectiveness after an appropriate learning curve and in carefully selected patients. Favourable elements include age over 20 years, competition level sports activity, recent instability, and absence of constitutional hyperlaxity.