Background: Flattening of the epiphyses of long bones is seen in several skeletal dysplasias and standardized measurements on a radiograph of the knee to detect skeletal dysplasias using this feature have been described. Since then only two other studies in which this method was used have been published, and both included only a small number of children and neither had a control group. In addition, the Dutch National Working Group on Skeletal Dysplasias began to have doubts about the reliability of the method. We therefore decided to re-evaluate its accuracy in a population of children with and without a skeletal dysplasia.
Objective: To determine the diagnostic value of standardized measurements on conventional AP radiographs of the knee in children with a skeletal dysplasia.
Subjects and methods: We measured the distal femoral metaphysis and epiphysis according to the published method on conventional AP radiographs of the knee in 45 healthy children and 52 children with a skeletal dysplasia. We compared graphically the height of the distal femoral epiphysis with its width and with the width of the femoral metaphysis. Receiver operating characteristic (ROC) curves were calculated for each group of children.
Results: All graphs showed a considerable overlap between children with a skeletal dysplasia and healthy children. The size of the area under the ROC curves for the different groups was small, varying between 0.567 and 0.653.
Conclusions: This method does not discriminate between children with a skeletal dysplasia and healthy children. We therefore consider it to be of little diagnostic value.