Comparison of two and three-dimensional computerized polyethylene wear analysis after total hip arthroplasty

Background: The accurate determination of acetabular polyethylene wear in vivo is necessary to assess the clinical performance of the bearing surfaces of total hip replacements. Our objective in this study was to determine the clinical performance of two and three-dimensional computerized wear analysis and to assess the implications of this performance on requirements for patient enrollment in studies designed to detect wear of total hip prostheses. Methods: Two and three-dimensional digital computerized analyses of acetabular polyethylene wear were performed on 153 hips in 140 patients. The acetabular components consisted of a polyethylene insert in a titanium shell, articulating with a 28-mm cobalt-chromium femoral component. The average duration of radiographic follow-up was 8.4 years. The correlation coefficient for two-dimensional versus three-dimensional analysis was calculated, as was the difference between the wear detected by the two techniques. The same observer analyzed each image twice, allowing an assessment of the repeatability of the two-dimensional and three-dimensional analyses. The impact of the clinical performance of each technique on the sample size needed for adequate power in prospective studies was evaluated. Results: There was a high correlation between two-dimensional and three-dimensional wear analysis (r(2) = 0.933). In thirty-one (5.2%) of 595 observations, the wear values derived with the two-dimensional and three-dimensional techniques were not consistent. Logistic regression demonstrated that acetabular anteversion had a significant effect on the likelihood of such inconsistency occurring. The two-dimensional technique detected 90.1% of the total linear wear subsequently detected by the three-dimensional analysis. The average wear value was 1.09 mm as detected by two-dimensional analysis and 1.21 mm as detected by three-dimensional analysis. The two-dimensional technique was four times more repeatable than the three-dimensional technique. Power analysis indicated that up to 1.4 times more patients need to be enrolled if the three-dimensional technique is used for wear analysis. Conclusions: While three-dimensional analysis detected 10% more wear, its repeatability was four times worse than that of the two-dimensional technique and, as a consequence, patient enrollment requirements for wear detection were higher. The poor quality of the lateral radiographs contributed to the decrease in the repeatability of the three-dimensional analysis. Three-dimensional analysis may be useful for highly anteverted cups, but the limited improvement in wear detection achieved with that technique, coupled with its inferior repeatability, limits its clinical value. Level of Evidence: Diagnostic study, Level IV-2 (poor reference standard). See Instructions to Authors for a complete description of levels of evidence.