259. RELIABILITY OF WEAR MEASUREMENTS FOR THE DUAL MOBILITY CUP: EXPERIMENTAL RADIOSTEREOMETRIC ANALYSIS (RSA)

Abstract

Purpose of the study: The concept introduced by Gilles Bousquet is an effective arm against dislocation of total hip arthroplasty (THA), as has been demonstrated in clinical series with a long follow-up. There remain certain questions concerning wear of dual mobility cups. We propose a radiostereometrical analysis (RSA) of femoral head migration in this type of implant. Our objective was to establish an accurate measurement and determine the intra- and interobserver variabilities.

Material and methods: A THA model was implanted and loaded with a simulator. Penetration of the implants was measured using a specially designed polyethylene insert with increasingly concentric wear (from 0, 0.25, 0.5 to 0.75 mm). Three investigators analysed (7 times in a double-blind protocol) the RSA images of these four inserts. The investigators were an expert (I), well-trained (II), naive (III). The accuracy of the measurement as well as the intra- and interobserver variabilities were determined using the root mean square (RMS) method, the interclass coefficient of correlation (ICC), the Bland and Altman analyses, and weighted Kappa analysis.

Results: Regarding accuracy, the RMS was 0.0388 [CI95: 0.02266–0.05564]. The mean error for preworn inserts was respectively 0.022mm (for 0.25mm prewear), 0.59mm (for 0.5mm), and 0.022mm (for 0.75mm). The intra-observer ICC was 0.9714 [0.9028–0.9918] for investigator I. The interobserver ICCs between investigators I and II and between I and III were respectively 0.943 and 0.968. The weighted kappa coefficients between I and II and between I and III were 0.827 and 0.849. The Bland and Altman analysis confirmed these results.

Discussion: Several RSA protocols could be designed to measure wear of the dual mobility cup. We chose detection of the wear pattern instead of the tantalum beads method. Our protocol, using a simple geometric model and not the manufacturers CAD files, showed an accuracy comparable with manufacturing tolerances with low variability.

Conclusion: This study validated our measurement method, a prerequisite for a randomized multicentric study which has been initiated to compare, by RSA, penetration of the head into the double mobility insert versus a fixed insert.