Between January 1987 and December 1988 there were 7575 births in the Swansea maternity unit. Of these 823 (10.9%) were considered to be at ‘high risk’ for developmental dysplasia of the hip (DDH). Static ultrasound examination was performed in each case and the results classified on the basis of the method of Graf. A total of 117 type III–IV hips in 83 infants was splinted using the Aberdeen splint.

Radiographs of these hips were taken at six and 12 months. Hilgenreiner’s measurements of the acetabular angle were made in all cases and the development of the femoral capital epiphysis was assessed by measuring the epiphyseal area. The effect of splintage on the acetabular angle and the epiphyseal area between the normal and abnormal splinted hips was compared. Radiographs of 16 normal infants (32 normal unsplinted hips) were used as a control group.

This cohort has now been followed up for a minimum of nine years. There have been no complications as a result of splintage. The failure rate was 1.7% or 0.25 per 1000 live births. No statistical difference was found when comparing the effect of splintage on the acetabular angle and epiphyseal area between normal and abnormal splinted hips and normal unsplinted hips.

Our study has shown that while the Aberdeen splint had a definite but small failure rate, it was safe in that it did not produce avascular necrosis. The current conventional view that a low rate of splintage is always best is therefore brought into question if the Aberdeen splint is chosen for the management of neonatal DDH.

The significance of cobalt as a cause of symptoms after McKee hip arthroplasty is discussed. Seven patients are described in whom such arthroplasties became unsatisfactory after periods varying from nine months to four years. Six of these patients were cobalt-positive but nickel-and chrome-negative on patch testing. Macroscopic and histological necrosis of bone, muscle and joint capsule around the prostheses was found in five patients whose hips were explored. The symptoms were progressive pain, a feeling of instability, and in two cases spontaneous dislocation. Radiological features included acetabular fracture, bone resorption, loosening and dislocation of the prosthesis. Increased cobalt concentrations (determined by atomic absorption spectrophotometry) in the urine of four patients and in a variety of tissues in one patient are presented. Patch testing is recommended in the investigation of patients with troublesome McKee hip arthroplasties.