The radiographic analysis of over 5000 metal on metal (MoM) hips using Ein Bild Roentgen Analyse (EBRA) software have been recently published in an attempt to determine the influence of cup orientation on bearing function. The validation of this software relies one study, conducted in a phantom pelvis without a femoral head in situ. Three dimensional computed tomographic (3D-CT) has been shown to be more accurate for hip and knee arthroplasty than plain radiographs for measurements of component orientation and position. The accuracy of EBRA when compared to 3D-CT for MoM hips specifically is unclear.

We measured the cup orientation of 96 large diameter MoM hips using EBRA analysis of plain radiographs and compared this to 3D-CT. All measurements were made using the radiographic definition of cup orientation. The mean difference in version between the two imaging modalities was 8°; with wide limits of agreement of −21.2° and +5.6°. Three retroverted cups were not detected by EBRA. The mean difference in inclination values was 0.3°, but there was up to 9°difference between imaging modalities. When measured by 3D CT, 64% of hips were within a 10° safe zone around 45° inclination and 20° version, compared to only 24% when measured by EBRA (Fishers Exact test, p< 0.0001).

The measurement of cup orientation of MoM hips using EBRA software is insufficiently accurate, particularly for the assessment of cup version. The cup rim is obscured by the large diameter femoral head on plain radiographs. Research studies using EBRA analysis for version have limited value if accuracy of more than 20 degrees is required to draw conclusions. This software may not be suitable to measure the performance of a device or surgeon. The limitations of EBRA can be overcome, if 3D-CT with an extended Hounsfield scale for data capture is used.