After many years of research into the detection and management of neonatal hip dysplasia there is a growing consensus that ultrasound is useful in the selection of hips which require splintage. It can detect some dysplastic hips which are missed on clinical examination 1-3 and exclude significant dysplasia in some hips which are thought to be abnormal clinically. 4 There is, however, relatively little information in the published literature on the use of ultrasound to assess the outcome and compare splinted hips with those without splintage and the effectiveness of different splints. As a result a variety of protocols for splintage is in use. In our hospital babies presenting with clinical instability of the hip may be treated by observation and no splintage or splinted using either a Pavlik harness, a Craig splint or a von Rosen splint. The management is determined by the orthopaedic consultant in charge of the patient, and this depends on which day of the week that the request for an opinion is received. Ultrasound assessment of babies with clinically suspected instability of the hip began in our hospital in 1993 and thereafter most hips were morphologically classified using the classification of Graf 5 before being seen in the orthopaedic clinic. Their progress is monitored by ultrasound. The radiologist (AGW) noted that dysplasia in babies treated with the von Rosen splint appeared to resolve more quickly than that which had other protocols of treatment. We have therefore reviewed the imaging and clinical outcome of all babies presenting with Graf type-III and type-IV dysplasia in order to assess the effectiveness of the different protocols of management.
Patients and MethodsThe ultrasound reports of all babies who were imaged for clinically suspected neonatal dysplasia of the hip between 1993 and 1998 were reviewed and those classified as Graf type III or type IV formed the basis of the study. Babies had been referred because of clinically suspected instability or the presence of risk factors for hip dysplasia. The scans were reassessed and those in which the quality of the image was inadequate (incorrect plane of imaging), or were not unequivocally Graf type III or type IV, were excluded from the study. Babies who were first imaged over the age of three months and those with a neurological abnormality were also excluded. A total of 134 hips in 96 babies (84 girls and 12 boys) was included in the study.The imaging was reviewed with particular attention to the first ultrasound examination, that carried out at 12 to 20