Simple mechanical devices did not improve pelvis positioning in AP pelvis radiographs for reliable assessment of the acetabular orientation

Abstract

Background: The aim of this study was to develop two simple positioning devices for anteroposterior pelvis radiographs and to evaluate their effect on accuracy of the radiographs for assessment of the acetabular orientation compared with non-instrumented positioning. Methods: The superior anterior iliac spines and the pubic symphysis were used as anatomical landmarks to obtain a horizontal orientation of the pelvis according to the anterior pelvic plane. Anteroposterior pelvis radiographs were taken of 11 human cadaveric pelvic bones with each of the positioning devices and without any device. Defined measurements were carried out to objectify the tilt and rotation of the pelvis and to assess the cross-over sign as well as the presence of the ischial spine sign. Computed tomography scans were performed as a standard of reference. Bland-Altman-Plots were used to compare the continuous measurement values and Cohen’s Kappa was applied for the categorical data. Intra- and inter-observer reliability was determined by the intraclass correlation coefficient and Cohen’s Kappa. Results: The mean values of the measurements showed a high variability. A low correlation of the measurement values was found between the radiographs of the different positioning methods and the computed tomography scans. The intra- and inter-observer reliability was good (Cohen’s Kappa 0.78 and 0.69; intraclass correlation coefficient 0.99 and 0.98). Conclusion: The use of positioning devices did not lead to more accurate anteroposterior pelvis radiographs compared to non-instrumented positioning. Simple positioning devices do not provide standardized anteroposterior pelvis radiographs for reliable assessment of the acetabular orientation.