Purpose of the study: Revision shoulder arthroplasty is generally considered to be a difficult procedure yielding modest improvement.

Material and methods: We report a prospective study of 45 patients, aged 69.8 years (range 49–85 years). Thirty-two patients had a simple humeral prosthesis and thirteen a total prosthesis. A reversed prosthesis was used for all revisions. The reasons for the revisions were classified into five groups: failure of prosthesis implanted for fracture (36%), glenoid problems of a total shoulder arthroplasty (24%), prosthetic instability (18%), failure of a hemiarthroplasty implanted for rotator cuff tear (11%), failure of a hemiarhtroplasty implanted for post-traumatic osteoarthritis (11%). The revision consisted in replacement with a reversed prosthesis. Patients were assessed pre and postoperatively using the Constant score for the clinical assessment and plain x-rays for the radiological assessment.

Results: Forty-one patients were reviewed at mean follow-up of 42.1 months (range 24–92). The four other patients died during the first two postoperative years. Subjectively, 73% of patients were satisfied. The Constant score improved from 187.7 to 55.6 on average. The best gain was obtained for the pain and daily activities scores.

Discussion: Revision shoulder arthroplasty provides only moderate improvement. Neer called a limited goal surgery. Results published on revision shoulder arthroplasty using a non-constrained prosthesis show that the functional gain is moderate. Revisio with a reversed total prosthesis gives better results because of the lesser impact of the cuff deficiency. The rate of complications after revision is greater than with first intention implantations.

Conclusion: Use of a reversed total shoulder prosthesis for revision shoulder arthroplasty provides encouraging results in terms of the mid-term functional outcome.

Purpose: We report our expeience with the medial and posteromedial approach to the humerus for plate fixation of fractures of the distal two-thirds of the humerus.

Material: Fifteen patients (eleven men and four women) were treated for fractures (n=13) or nonunion (n=2) situated below the proximal third of the humerus without radial nerve involvement.

Methods: Eight patients were installed in the supine position for median approach between the humerus bundle and the median nerve anteriorly and the ulnar nerve posteriorly. The posteriomedian approach, with the ulnar nerve posteriorly and the brachial triceps anteriorly, was used for seven other patients installed in the prone position. The fixation plate applied to the medial aspect allowed at least six corticals on either side of the fracture line. The patients were immobilised for 45 days. Passive rehabilitation exercises involved the elbow and the shoulder without external rotation. Clinical and x-ray follow-up data were available for all patients.

Results: One patient was lost to follow-up two months after surgery: at this time the x-ray had demonstrated bone healing. Function could not be assessed as the fracture had occurred on the same side as the hemiplegia also caused by the initial trauma. For the fourteen other patients, mean follow-up was 12 months (range 6 – 36). Three patients operated via the median approach presented paraesthesia in the median nerve territory which was regressive in two. There were no neurological complications in the posteromedian approach group. Function was good for elbow and shoulder except for two patients. Bone healing was achieved in all cases.

Discussion: These approaches allowed avoiding dissection of the radial nerve and provided a more aesthetic scar. Several difficulties were encountered with the median approach and reduction was difficult to control. In such cases it is advisable to widen the exposure to avoid stretching the median nerve. These approaches are contraindicated if radial nerve injury is identified preoperatively.

Conclusion: Osteosynthesis of the humerus via a median approach avoids the need for radial nerve dissection. Reduction appears to be easier via the posteromedian approach with less risk than with the median approach.

Purpose: It is known that severe distention of the inferior glenohumeral ligament (IGHL) during anteror-inferior shoulder instability is an important factor of poor functional prognosis after arthroscopic stabilisation. O. Gagey proposed a clinical test to assess laxity of the IGHL. The purpose of this study was to assess the laxity of the IGHL using a dynamic radiological test (AP view in passive abduction of the glenohu-meral joint) and to correlate findings with arthroscopic observations..

Material and methods: We performed a prospective study in 21 patients scheduled for arthroscopic stabilisation for anteroinferior shoulder instability. Mean age was 24.6 years, 17 men and 4 women. The test was performed in the supine position with a strictly AP view of the shoulder. Bilateral comparative images were obtained. The shoulder was brought to forced passive abduction in neutral rotation without general anaesthesia or locoregional anaesthesia. The angle between the axis of the humeral diaphysis and the line passing through the inferior border of the glenoid cavity and the lateral border of the scapular tubercle was measured. Vuillemin has demonstrated that this test is reliable and reproducible. During arthroscopy performed for diagnostic and therapeutic purposes, the degree of distension was quantified using the Detrisac classification of four stages. We considered that stages 3 and 4 were frank pathological distension. We used the threshold of 15° for the difference between the healthy and pathological side for the radiological test. We assessed the ability to demonstrate severe laxity of the IGHL.

Results: For differences in abduction less than 15°, the test sensitivity was 77%, specificity 91%, positive predictive value 87% and negative predictive value 84%.

Discussion: A careful physical examination and appropriate complementary tests are essential for the evaluation of anteroinferior instability of the shoulder joint in order to obtain a precise diagnosis and search for contraindications for arthroscopic cure. The rate of recurrence after arthroscopic stabilisation remains above that obtained with open techniques. It has been demonstrated that major laxity of the IGHL constitutes a relative contraindication for arthroscopic stabilisation. Radiographic measurements provide precise information for evaluating the laxity of the IGHL. Taking a positive threshold of 15° difference identifies 87% of the cases of Detrisac stage 3 or 4 ligament distension.

Conclusion: We propose a preoperative complementary test using standard x-rays together with our dynamic radiological test of passive shoulder abduction. If the difference between the healthy and pathological side is greater or equal to 15°, the therapeutic strategy should include not only reinsertion of the rim but also retight-ening the ligament complex, or open stabilisation.