Inter- and intra-observer variability of radiography and computed tomography for evaluation of Zurich cementless acetabular cup placement ex vivo.

Abstract

To evaluate the inter- and intra-observer variability in measurement of the angle of lateral opening (ALO) and version angle measurement using digital radiography and computed tomography (CT). Each hemipelvis was implanted with a cementless acetabular cup. Ventrodorsal and mediolateral radiographs were made of each pelvis, followed by CT imaging. After removal of the first cup, the pelves were implanted with an acetabular cup in the contralateral acetabulum and imaging was repeated. Three surgeons measured the ALO and version angles three times for each cup from the mediolateral radiographic projection. The same measurements were made using three-dimensional multiplanar reconstructions from CT images. Two anatomical axes were used to measure pelvic inclination in the sagittal plane, resulting in six measurements per cup. Two-way repeated measures analysis of variance evaluated inter- and intra-observer repeatability for radiographic and CT-based measurements. Version angle based on radiographic measurement did not differ within surgeons (p = 0.433), but differed between surgeons (p <0.001). Radiographic measurement of ALO differed within surgeons (p = 0.006) but not between surgeons (p = 0.989). The ALO and version angle measured on CT images did not differ with or between surgeons. Assessment of inter- and intra-observer measurement of ALO and version angle was more reproducible using CT images than conventional mediolateral radiography for a Zurich cementless acetabular cup.