2D measurements of cup orientation are less reliable than 3D measurements: A retrospective study of 87 metal-on-metal hips

BACKGROUND AND PURPOSE: 2D analysis of metal-on-metal (MoM) hip arthroplasty (HA) has been conducted in several large series on conventional radiographs with the use of Ein Bild Roentegen Analyse (EBRA) software, but there have been no comparisons with 3D analysis in the literature. The main aim of this study was to quantify the agreement in measurements of cup version of large-diameter MoM hips obtained by EBRA and by 3D computed tomography (3D-CT). The secondary aim was to quantify the agreement for cup inclination. Lastly, we wanted to determine the inter- and intra-observer reliability of both methods. PATIENTS AND METHODS: 87 MoM hips in 81 patients were analyzed for cup inclination and version in 2D on conventional radiographs using EBRA software. The results were compared with 3D measurements using CT. RESULTS: Cup version was underestimated by EBRA when compared to 3D-CT, by 6° on average with the pelvis supine and by 8° on average with the pelvis orientated to the anterior pelvic plane (APP). For inclination, the mean difference was no more than 1°. 53% of hips were within a 10° safe zone of 45° inclination and 20° version when measured by 3D-CT with the pelvis supine (and 54% with the pelvis in the APP). The proportion was only 24% when measured by EBRA. Inter- and intra-observer reliability of cup version is poorer using 2D analysis than when using 3D-CT. INTERPRETATION: Errors in version in 2D were due to the difficulty in delineating the cup rim, which was obscured by a large-diameter metal head of the same radio-opacity. This can be overcome with 3D analysis. The present study demonstrates that measurements using EBRA have poor agreement and are less reliable than those with 3D-CT when measuring cup version and inclination in MoM hips.