We report the 20-year results of Bonnin's modification of the Bristow-Latarjet procedure in 14 patients operated on by one surgeon. All but one patient had had traumatic dislocations. At review, the Rowe scores were excellent in five, good in eight and fair in one. The functional outcome was satisfactory, with a mean Constant-Murley score of 80 points (68 to 95), but 12 patients had restriction of external rotation (86%). There were radiological degenerative changes in ten shoulders (71%): six in Samilson grade I, one in grade II, and three in grade III. Isometric power was considerably reduced in patients with grade-III degenerative change. This operation provides good long-term shoulder stability, but the high incidence of radiological degenerative change is a cause for concern.

Sickle-cell disease (SCD) is probably the commonest cause of avascular necrosis worldwide, and its prevalence appears to be rising in developed countries. Avascular necrosis of the humeral head is a common complication but has not been previously studied in detail. We have reviewed 138 patients with SCD for clinical, radiological and functional abnormalities of the shoulder, using a radiological classification designed for avascular necrosis of the shoulder. Radiographic lesions, frequently bilateral, were found in 28% and only 53% of patients had normal shoulder function. The management of this relatively common complication is difficult. Joint replacement is likely to fail and early diagnosis is important.

A prospective, randomised and independently assessed trial of the Ring UPM total hip replacement showed that the quality of the early result was better if the femoral prosthesis was cemented than if it was not. More patients with cemented prostheses were painfree at four months (58% cemented:42% uncemented) and at one year (63% cemented:50% uncemented), but at two years pain relief was equal in both groups. At two years significantly more patients with cemented prostheses could walk without support (96% cemented:62% uncemented, p = 0.01 to 0.05). There is a need for more similar trials to compare the results of contemporary designs of cemented and uncemented total hip prostheses.

One hundred and fifty asymptomatic shoulders in 75 schoolchildren were studied. The shoulders were tested for instability and a hyperextensometer was used to assess joint laxity. Signs of instability were found in 57% of the shoulders in boys and 48% in girls; the commonest sign was a positive posterior drawer test which was found in 63 shoulders. A positive sulcus sign was found in 17 shoulders and 17 subjects had signs of multidirectional instability. General joint laxity was not a feature of subjects whose shoulders had positive instability signs.

We performed a randomised prospective trial to compare the results of 27 cemented and 26 uncemented bipolar hemiarthroplasties in active patients with displaced subcapital fractures of the femoral neck. After a mean follow-up of 17 months, significantly more of the uncemented group were experiencing pain in the hip and using more walking aids than the patients in the cemented group. The incidence of postoperative complications, the early mortality rate and the operating time and blood loss were not significantly different. Using otherwise identical prostheses the early results were much better with a cemented Thompson stem than with an uncemented Austin Moore stem.

We report the complications of prophylactic pinning of slipped upper femoral epiphysis with Crawford Adams pins in 95 cases. Complications of pin placement were seen in 13.7%. Although seven hips had penetration of the joint, there were no cases of chondrolysis or avascular necrosis. Excavation of the lateral femoral cortex was required at pin removal in 12.5% of cases. Analysis of the growth around pins allowed recommendations to be made regarding pin protrusion. The use of improved fixation devices may reduce the need for multiple pins.

Ten epiphyses in seven children underwent fixed-rate distraction of 0.25 mm twice daily in an attempt to achieve percutaneous leg lengthening by chondrodiatasis. The forces generated across the growth plate were recorded by means of strain gauges incorporated into the distractors. All epiphyses fractured before 33 days of lengthening. An average gain of 6.75 cm was achieved. Epiphyseal distraction at the lower femur produced many complications, but at the upper tibial epiphysis planned lengthening was achieved, with excellent bone production and few complications.

It has been shown that raised intracapsular pressure causes avascular necrosis of the femoral head in experimental animals, but the relevance of this to clinical fractures of the femoral neck is controversial. We have studied 19 patients with intracapsular fractures of the femoral neck by pressure measurement and by ultrasonography to demonstrate capsular distension. The intra-articular pressure in Garden Grade I and II fractures averaged 66.4 mmHg with a maximum of 145 mmHg. In 10 Garden Grade III and IV fractures the average pressure was 28 mmHg with a maximum of 65 mmHg. Most of the recorded intracapsular pressures were high enough to have caused possible vascular embarrassment, and it is suggested that early decompression of the haemarthrosis should be considered.