Can predictive and functional methods locate the dysplastic hip joint center?

Abstract

Background: Properly characterizing biomechanics in populations with known hip deformities, such as developmental dysplasia of the hip (DDH), requires accurately identifying the hip joint center (HJC). Many studies have assessed the accuracy or repeatability of predictive and functional methods to locate the HJC in normally shaped hips, but not for hips with DDH.

Research question: For patients with DDH, which of three common functional and four common predictive HJC methods is most accurate compared to a magnetic-resonance-imaging-based anatomic method?

Methods: HJC method accuracy was calculated as the resultant vector distance from the anatomic HJCs. Bias was calculated as distance errors along anatomical directions. Accuracy and bias were tested in patients with DDH (N = 30) and controls with normally shaped hips (N = 20) using each HJC method and two separate techniques for aligning the image-based and skin-marker-based pelvis segments.

Results: Using each alignment technique, the predictive Harrington (pelvis-only) method was the most accurate (smallest resultant errors) in more patients with DDH than any other method. The Harrington (pelvis-only) method also had the lowest or nearly lowest bias in each direction. For controls, the Harrington (pelvis-only) method also had among the lowest errors but there was not a standout "most accurate" method.

Significance: Although no method is perfect, in the absence of imaging data and an anatomic HJC, this study supports the predictive Harrington (pelvis-only) HJC as the most appropriate for dysplastic hips.