The management and outcome of treatment in 42 patients (49 shoulders) with an infected shoulder prosthesis was reviewed in a retrospective multicentre study of 2343 prostheses. The factors which were analysed included the primary diagnosis, the delay between the diagnosis of infection and treatment and the type of treatment. Treatment was considered to be successful in 30 patients (71%). Previous surgery and radiotherapy were identified as risk factors for the development of infection. All patients with an infected prosthesis had pain and limitation of movement and 88% showed radiological loosening. In 50% of the shoulders, the antibiotics chosen and the length of treatment were considered not to be optimal. The mean follow-up was 34 months. Antibiotics or debridement alone were ineffective. In acute infection, immediate revision with excision of all infected tissue and exchange of the prosthesis with appropriate antibiotic therapy gave the best results. Multidisciplinary collaboration is recommended

We studied 100 fresh human shoulders in cadavers (mean age 76 years), and the range of passive abduction (RPA) in 100 volunteers with normal shoulders and in 90 patients with instability of the joint, over a period of six years. The anatomical and clinical findings showed that passive abduction occurs within the glenohumeral joint only, is controlled by the inferior glenohumeral ligament and has a constant value in 95% of both shoulders in normal subjects. In patients with instability, 85% showed an RPA of over 105° with 90° in the contralateral shoulder. In the remaining patients a strongly positive apprehension test suggested a diagnosis of instability. An RPA of more than 105° is associated with lengthening and laxity of the inferior glenohumeral ligament