RADIOGRAPHIC EVALUATION OF POLYETHYLENE WEAR IN ACETABULAR COMPONENTS: COMPARISON OF 2D AND 3D MEASUREMENT TECHNIQUES WITH ASSESSMENT OF INTEROBSERVER VARIATION

Abstract

The purpose of this study was to compare the 2D and 3D linear and volume wear readings of the three most commonly used methods for measuring polyethylene wear: the Livermore, Devane and Martell techniques. Inter-observer variation of measurements using the techniques of Devane and of Martell on conventional radiographs was also performed. The radiographs of 80 patients (mean age 60+/−10 yrs) who had a Harris-Galante I total hip arthroplasty were measured. Nine different reviewers for the Devane technique readings including Dr Devane and eight reviewers for the Martell technique readings including Dr Martell made blinded independent wear observations for each radiograph set. One reviewer measured the 20 annual linear wear rate for all radiograph sets using the 2D Livermore technique. Inter-observer variation as a function of patient, reviewer, and total variation was statistically assessed using variance component analysis. Mean wear measured using the Livermore technique was the same as with the Devane and Martell method, but with a greater variation. Comparison of the Devane and Martell method for patient STD, reviewer STD, error STD (multiple reviews of same radiographs), total STD (randomly picked reviewer), mostly show a mean 50% lower STD with the Devane technique. Correlation (correlation coefficient of two randomly selected reviewers) is significantly better with the Devane technique.