Measurement of Pubofemoral Distance in the Diagnosis of Developmental Dysplasia of the Hip: Sensitivity and Specificity.

Abstract

To evaluate the accuracy of measurement of the pubofemoral distance (PFD) for the diagnosis of developmental dysplasia of the hip (DDH), using the Graf method as the reference standard. This was a prospective diagnostic accuracy study evaluating 1980 hips in at-risk neonates. The PFD measurement and the Graf method were performed at the same opportunity, with the hips in the same position (slightly flexed, adducted, and medially rotated). In our sample, the prevalence of DDH, defined as Graf type IIb or higher, was 15.6%. The mean PFD for dysplastic hips was 3.5 mm (median, 3.3 mm), whereas it was 3.0 mm (median, 2.9 mm) for nondysplastic hips. In both groups, there was a significant overlap between the PFD measurements and the Graf classifications. The PFD cutoff with the largest area under the receiver operating characteristic curve was 3.0 mm, which was found to have a sensitivity and specificity of 63.6 and 62.2%, respectively, with positive and negative predictive values of 31.2 and 71.4%, respectively, for the diagnosis of DDH. Measurement of the PFD shows good overall accuracy for the diagnosis of DDH. However, given its low-positive predictive value, it should not be considered to be a replacement for the Graf method.